Classification - Machine Learning This is ‘Classification’ tutorial which is a part of the Machine Learning course offered by Simplilearn. We will learn Classification algorithms, types of classification algorithms, support vector machines(SVM), Naive Bayes, Decision Tree and Random Forest Classifier in this tutorial. Objectives Let us look at some of the objectives covered under this section of Machine Learning tutorial. Define Classification and list its algorithms Describe Logistic Regression and Sigmoid Probability Explain K-Nearest Neighbors and KNN classification Understand Support Vector Machines, Polynomial Kernel, and Kernel Trick Analyze Kernel Support Vector Machines with an example Implement the Naïve Bayes Classifier Demonstrate Decision Tree Classifier Describe Random Forest Classifier Classification: Meaning Classification is a type of supervised learning. It specifies the class to which data elements belong to and is best used when the output has finite and discrete values. It predicts a class for an input variable as well. There are 2 types of Classification: Binomial Multi-Class Classification: Use Cases Some of the key areas where classification cases are being used: To find whether an email received is a spam or ham To identify customer segments To find if a bank loan is granted To identify if a kid will pass or fail in an examination Classification: Example Social media sentiment analysis has two potential outcomes, positive or negative, as displayed by the chart given below. https://www.simplilearn.com/ice9/free_resources_article_thumb/classification-example-machine-learning.JPG This chart shows the classification of the Iris flower dataset into its three sub-species indicated by codes 0, 1, and 2. https://www.simplilearn.com/ice9/free_resources_article_thumb/iris-flower-dataset-graph.JPG The test set dots represent the assignment of new test data points to one class or the other based on the trained classifier model. Types of Classification Algorithms Let’s have a quick look into the types of Classification Algorithm below. Linear Models Logistic Regression Support Vector Machines Nonlinear models K-nearest Neighbors (KNN) Kernel Support Vector Machines (SVM) Naïve Bayes Decision Tree Classification Random Forest Classification Logistic Regression: Meaning Let us understand the Logistic Regression model below. This refers to a regression model that is used for classification. This method is widely used for binary classification problems. It can also be extended to multi-class classification problems. Here, the dependent variable is categorical: y ϵ {0, 1} A binary dependent variable can have only two values, like 0 or 1, win or lose, pass or fail, healthy or sick, etc In this case, you model the probability distribution of output y as 1 or 0. This is called the sigmoid probability (σ). If σ(θ Tx) > 0.5, set y = 1, else set y = 0 Unlike Linear Regression (and its Normal Equation solution), there is no closed form solution for finding optimal weights of Logistic Regression. Instead, you must solve this with maximum likelihood estimation (a probability model to detect the maximum likelihood of something happening). It can be used to calculate the probability of a given outcome in a binary model, like the probability of being classified as sick or passing an exam. https://www.simplilearn.com/ice9/free_resources_article_thumb/logistic-regression-example-graph.JPG Sigmoid Probability The probability in the logistic regression is often represented by the Sigmoid function (also called the logistic function or the S-curve): https://www.simplilearn.com/ice9/free_resources_article_thumb/sigmoid-function-machine-learning.JPG In this equation, t represents data values * the number of hours studied and S(t) represents the probability of passing the exam. Assume sigmoid function: https://www.simplilearn.com/ice9/free_resources_article_thumb/sigmoid-probability-machine-learning.JPG g(z) tends toward 1 as z -> infinity , and g(z) tends toward 0 as z -> infinity K-nearest Neighbors (KNN) K-nearest Neighbors algorithm is used to assign a data point to clusters based on similarity measurement. It uses a supervised method for classification. The steps to writing a k-means algorithm are as given below: https://www.simplilearn.com/ice9/free_resources_article_thumb/knn-distribution-graph-machine-learning.JPG Choose the number of k and a distance metric. (k = 5 is common) Find k-nearest neighbors of the sample that you want to classify Assign the class label by majority vote. KNN Classification A new input point is classified in the category such that it has the most number of neighbors from that category. For example: https://www.simplilearn.com/ice9/free_resources_article_thumb/knn-classification-machine-learning.JPG Classify a patient as high risk or low risk. Mark email as spam or ham. Keen on learning about Classification Algorithms in Machine Learning? Click here! Support Vector Machine (SVM) Let us understand Support Vector Machine (SVM) in detail below. SVMs are classification algorithms used to assign data to various classes. They involve detecting hyperplanes which segregate data into classes. SVMs are very versatile and are also capable of performing linear or nonlinear classification, regression, and outlier detection. Once ideal hyperplanes are discovered, new data points can be easily classified. https://www.simplilearn.com/ice9/free_resources_article_thumb/support-vector-machines-graph-machine-learning.JPG The optimization objective is to find “maximum margin hyperplane” that is farthest from the closest points in the two classes (these points are called support vectors). In the given figure, the middle line represents the hyperplane. SVM Example Let’s look at this image below and have an idea about SVM in general. Hyperplanes with larger margins have lower generalization error. The positive and negative hyperplanes are represented by: https://www.simplilearn.com/ice9/free_resources_article_thumb/positive-negative-hyperplanes-machine-learning.JPG Classification of any new input sample xtest : If w0 + wTxtest > 1, the sample xtest is said to be in the class toward the right of the positive hyperplane. If w0 + wTxtest < -1, the sample xtest is said to be in the class toward the left of the negative hyperplane. When you subtract the two equations, you get: https://www.simplilearn.com/ice9/free_resources_article_thumb/equation-subtraction-machine-learning.JPG Length of vector w is (L2 norm length): https://www.simplilearn.com/ice9/free_resources_article_thumb/length-of-vector-machine-learning.JPG You normalize with the length of w to arrive at: https://www.simplilearn.com/ice9/free_resources_article_thumb/normalize-equation-machine-learning.JPG SVM: Hard Margin Classification Given below are some points to understand Hard Margin Classification. The left side of equation SVM-1 given above can be interpreted as the distance between the positive (+ve) and negative (-ve) hyperplanes; in other words, it is the margin that can be maximized. Hence the objective of the function is to maximize with the constraint that the samples are classified correctly, which is represented as : https://www.simplilearn.com/ice9/free_resources_article_thumb/hard-margin-classification-machine-learning.JPG This means that you are minimizing ‖w‖. This also means that all positive samples are on one side of the positive hyperplane and all negative samples are on the other side of the negative hyperplane. This can be written concisely as : https://www.simplilearn.com/ice9/free_resources_article_thumb/hard-margin-classification-formula.JPG Minimizing ‖w‖ is the same as minimizing. This figure is better as it is differentiable even at w = 0. The approach listed above is called “hard margin linear SVM classifier.” SVM: Soft Margin Classification Given below are some points to understand Soft Margin Classification. To allow for linear constraints to be relaxed for nonlinearly separable data, a slack variable is introduced. (i) measures how much ith instance is allowed to violate the margin. The slack variable is simply added to the linear constraints. https://www.simplilearn.com/ice9/free_resources_article_thumb/soft-margin-calculation-machine-learning.JPG Subject to the above constraints, the new objective to be minimized becomes: https://www.simplilearn.com/ice9/free_resources_article_thumb/soft-margin-calculation-formula.JPG You have two conflicting objectives now—minimizing slack variable to reduce margin violations and minimizing to increase the margin. The hyperparameter C allows us to define this trade-off. Large values of C correspond to larger error penalties (so smaller margins), whereas smaller values of C allow for higher misclassification errors and larger margins. https://www.simplilearn.com/ice9/free_resources_article_thumb/machine-learning-certification-video-preview.jpg SVM: Regularization The concept of C is the reverse of regularization. Higher C means lower regularization, which increases bias and lowers the variance (causing overfitting). https://www.simplilearn.com/ice9/free_resources_article_thumb/concept-of-c-graph-machine-learning.JPG IRIS Data Set The Iris dataset contains measurements of 150 IRIS flowers from three different species: Setosa Versicolor Viriginica Each row represents one sample. Flower measurements in centimeters are stored as columns. These are called features. IRIS Data Set: SVM Let’s train an SVM model using sci-kit-learn for the Iris dataset: https://www.simplilearn.com/ice9/free_resources_article_thumb/svm-model-graph-machine-learning.JPG Nonlinear SVM Classification There are two ways to solve nonlinear SVMs: by adding polynomial features by adding similarity features Polynomial features can be added to datasets; in some cases, this can create a linearly separable dataset. https://www.simplilearn.com/ice9/free_resources_article_thumb/nonlinear-classification-svm-machine-learning.JPG In the figure on the left, there is only 1 feature x1. This dataset is not linearly separable. If you add x2 = (x1)2 (figure on the right), the data becomes linearly separable. Polynomial Kernel In sci-kit-learn, one can use a Pipeline class for creating polynomial features. Classification results for the Moons dataset are shown in the figure. https://www.simplilearn.com/ice9/free_resources_article_thumb/polynomial-kernel-machine-learning.JPG Polynomial Kernel with Kernel Trick Let us look at the image below and understand Kernel Trick in detail. https://www.simplilearn.com/ice9/free_resources_article_thumb/polynomial-kernel-with-kernel-trick.JPG For large dimensional datasets, adding too many polynomial features can slow down the model. You can apply a kernel trick with the effect of polynomial features without actually adding them. The code is shown (SVC class) below trains an SVM classifier using a 3rd-degree polynomial kernel but with a kernel trick. https://www.simplilearn.com/ice9/free_resources_article_thumb/polynomial-kernel-equation-machine-learning.JPG The hyperparameter coefθ controls the influence of high-degree polynomials. Kernel SVM Let us understand in detail about Kernel SVM. Kernel SVMs are used for classification of nonlinear data. In the chart, nonlinear data is projected into a higher dimensional space via a mapping function where it becomes linearly separable. https://www.simplilearn.com/ice9/free_resources_article_thumb/kernel-svm-machine-learning.JPG In the higher dimension, a linear separating hyperplane can be derived and used for classification. A reverse projection of the higher dimension back to original feature space takes it back to nonlinear shape. As mentioned previously, SVMs can be kernelized to solve nonlinear classification problems. You can create a sample dataset for XOR gate (nonlinear problem) from NumPy. 100 samples will be assigned the class sample 1, and 100 samples will be assigned the class label -1. https://www.simplilearn.com/ice9/free_resources_article_thumb/kernel-svm-graph-machine-learning.JPG As you can see, this data is not linearly separable. https://www.simplilearn.com/ice9/free_resources_article_thumb/kernel-svm-non-separable.JPG You now use the kernel trick to classify XOR dataset created earlier. https://www.simplilearn.com/ice9/free_resources_article_thumb/kernel-svm-xor-machine-learning.JPG Naïve Bayes Classifier What is Naive Bayes Classifier? Have you ever wondered how your mail provider implements spam filtering or how online news channels perform news text classification or even how companies perform sentiment analysis of their audience on social media? All of this and more are done through a machine learning algorithm called Naive Bayes Classifier. Naive Bayes Named after Thomas Bayes from the 1700s who first coined this in the Western literature. Naive Bayes classifier works on the principle of conditional probability as given by the Bayes theorem. Advantages of Naive Bayes Classifier Listed below are six benefits of Naive Bayes Classifier. Very simple and easy to implement Needs less training data Handles both continuous and discrete data Highly scalable with the number of predictors and data points As it is fast, it can be used in real-time predictions Not sensitive to irrelevant features Bayes Theorem We will understand Bayes Theorem in detail from the points mentioned below. According to the Bayes model, the conditional probability P(Y|X) can be calculated as: P(Y|X) = P(X|Y)P(Y) / P(X) This means you have to estimate a very large number of P(X|Y) probabilities for a relatively small vector space X. For example, for a Boolean Y and 30 possible Boolean attributes in the X vector, you will have to estimate 3 billion probabilities P(X|Y). To make it practical, a Naïve Bayes classifier is used, which assumes conditional independence of P(X) to each other, with a given value of Y. This reduces the number of probability estimates to 2*30=60 in the above example. Naïve Bayes Classifier for SMS Spam Detection Consider a labeled SMS database having 5574 messages. It has messages as given below: https://www.simplilearn.com/ice9/free_resources_article_thumb/naive-bayes-spam-machine-learning.JPG Each message is marked as spam or ham in the data set. Let’s train a model with Naïve Bayes algorithm to detect spam from ham. The message lengths and their frequency (in the training dataset) are as shown below: https://www.simplilearn.com/ice9/free_resources_article_thumb/naive-bayes-spam-spam-detection.JPG Analyze the logic you use to train an algorithm to detect spam: Split each message into individual words/tokens (bag of words). Lemmatize the data (each word takes its base form, like “walking” or “walked” is replaced with “walk”). Convert data to vectors using scikit-learn module CountVectorizer. Run TFIDF to remove common words like “is,” “are,” “and.” Now apply scikit-learn module for Naïve Bayes MultinomialNB to get the Spam Detector. This spam detector can then be used to classify a random new message as spam or ham. Next, the accuracy of the spam detector is checked using the Confusion Matrix. For the SMS spam example above, the confusion matrix is shown on the right. Accuracy Rate = Correct / Total = (4827 + 592)/5574 = 97.21% Error Rate = Wrong / Total = (155 + 0)/5574 = 2.78% https://www.simplilearn.com/ice9/free_resources_article_thumb/confusion-matrix-machine-learning.JPG Although confusion Matrix is useful, some more precise metrics are provided by Precision and Recall. https://www.simplilearn.com/ice9/free_resources_article_thumb/precision-recall-matrix-machine-learning.JPG Precision refers to the accuracy of positive predictions. https://www.simplilearn.com/ice9/free_resources_article_thumb/precision-formula-machine-learning.JPG Recall refers to the ratio of positive instances that are correctly detected by the classifier (also known as True positive rate or TPR). https://www.simplilearn.com/ice9/free_resources_article_thumb/recall-formula-machine-learning.JPG Precision/Recall Trade-off To detect age-appropriate videos for kids, you need high precision (low recall) to ensure that only safe videos make the cut (even though a few safe videos may be left out). The high recall is needed (low precision is acceptable) in-store surveillance to catch shoplifters; a few false alarms are acceptable, but all shoplifters must be caught. Learn about Naive Bayes in detail. Click here! Decision Tree Classifier Some aspects of the Decision Tree Classifier mentioned below are. Decision Trees (DT) can be used both for classification and regression. The advantage of decision trees is that they require very little data preparation. They do not require feature scaling or centering at all. They are also the fundamental components of Random Forests, one of the most powerful ML algorithms. Unlike Random Forests and Neural Networks (which do black-box modeling), Decision Trees are white box models, which means that inner workings of these models are clearly understood. In the case of classification, the data is segregated based on a series of questions. Any new data point is assigned to the selected leaf node. https://www.simplilearn.com/ice9/free_resources_article_thumb/decision-tree-classifier-machine-learning.JPG Start at the tree root and split the data on the feature using the decision algorithm, resulting in the largest information gain (IG). This splitting procedure is then repeated in an iterative process at each child node until the leaves are pure. This means that the samples at each node belonging to the same class. In practice, you can set a limit on the depth of the tree to prevent overfitting. The purity is compromised here as the final leaves may still have some impurity. The figure shows the classification of the Iris dataset. https://www.simplilearn.com/ice9/free_resources_article_thumb/decision-tree-classifier-graph.JPG IRIS Decision Tree Let’s build a Decision Tree using scikit-learn for the Iris flower dataset and also visualize it using export_graphviz API. https://www.simplilearn.com/ice9/free_resources_article_thumb/iris-decision-tree-machine-learning.JPG The output of export_graphviz can be converted into png format: https://www.simplilearn.com/ice9/free_resources_article_thumb/iris-decision-tree-output.JPG Sample attribute stands for the number of training instances the node applies to. Value attribute stands for the number of training instances of each class the node applies to. Gini impurity measures the node’s impurity. A node is “pure” (gini=0) if all training instances it applies to belong to the same class. https://www.simplilearn.com/ice9/free_resources_article_thumb/impurity-formula-machine-learning.JPG For example, for Versicolor (green color node), the Gini is 1-(0/54)2 -(49/54)2 -(5/54) 2 ≈ 0.168 https://www.simplilearn.com/ice9/free_resources_article_thumb/iris-decision-tree-sample.JPG Decision Boundaries Let us learn to create decision boundaries below. For the first node (depth 0), the solid line splits the data (Iris-Setosa on left). Gini is 0 for Setosa node, so no further split is possible. The second node (depth 1) splits the data into Versicolor and Virginica. If max_depth were set as 3, a third split would happen (vertical dotted line). https://www.simplilearn.com/ice9/free_resources_article_thumb/decision-tree-boundaries.JPG For a sample with petal length 5 cm and petal width 1.5 cm, the tree traverses to depth 2 left node, so the probability predictions for this sample are 0% for Iris-Setosa (0/54), 90.7% for Iris-Versicolor (49/54), and 9.3% for Iris-Virginica (5/54) CART Training Algorithm Scikit-learn uses Classification and Regression Trees (CART) algorithm to train Decision Trees. CART algorithm: Split the data into two subsets using a single feature k and threshold tk (example, petal length < “2.45 cm”). This is done recursively for each node. k and tk are chosen such that they produce the purest subsets (weighted by their size). The objective is to minimize the cost function as given below: https://www.simplilearn.com/ice9/free_resources_article_thumb/cart-training-algorithm-machine-learning.JPG The algorithm stops executing if one of the following situations occurs: max_depth is reached No further splits are found for each node Other hyperparameters may be used to stop the tree: min_samples_split min_samples_leaf min_weight_fraction_leaf max_leaf_nodes Gini Impurity or Entropy Entropy is one more measure of impurity and can be used in place of Gini. https://www.simplilearn.com/ice9/free_resources_article_thumb/gini-impurity-entrophy.JPG It is a degree of uncertainty, and Information Gain is the reduction that occurs in entropy as one traverses down the tree. Entropy is zero for a DT node when the node contains instances of only one class. Entropy for depth 2 left node in the example given above is: https://www.simplilearn.com/ice9/free_resources_article_thumb/entrophy-for-depth-2.JPG Gini and Entropy both lead to similar trees. DT: Regularization The following figure shows two decision trees on the moons dataset. https://www.simplilearn.com/ice9/free_resources_article_thumb/dt-regularization-machine-learning.JPG The decision tree on the right is restricted by min_samples_leaf = 4. The model on the left is overfitting, while the model on the right generalizes better. Random Forest Classifier Let us have an understanding of Random Forest Classifier below. A random forest can be considered an ensemble of decision trees (Ensemble learning). Random Forest algorithm: Draw a random bootstrap sample of size n (randomly choose n samples from the training set). Grow a decision tree from the bootstrap sample. At each node, randomly select d features. Split the node using the feature that provides the best split according to the objective function, for instance by maximizing the information gain. Repeat the steps 1 to 2 k times. (k is the number of trees you want to create, using a subset of samples) Aggregate the prediction by each tree for a new data point to assign the class label by majority vote (pick the group selected by the most number of trees and assign new data point to that group). Random Forests are opaque, which means it is difficult to visualize their inner workings. https://www.simplilearn.com/ice9/free_resources_article_thumb/random-forest-classifier-graph.JPG However, the advantages outweigh their limitations since you do not have to worry about hyperparameters except k, which stands for the number of decision trees to be created from a subset of samples. RF is quite robust to noise from the individual decision trees. Hence, you need not prune individual decision trees. The larger the number of decision trees, the more accurate the Random Forest prediction is. (This, however, comes with higher computation cost). Key Takeaways Let us quickly run through what we have learned so far in this Classification tutorial. Classification algorithms are supervised learning methods to split data into classes. They can work on Linear Data as well as Nonlinear Data. Logistic Regression can classify data based on weighted parameters and sigmoid conversion to calculate the probability of classes. K-nearest Neighbors (KNN) algorithm uses similar features to classify data. Support Vector Machines (SVMs) classify data by detecting the maximum margin hyperplane between data classes. Naïve Bayes, a simplified Bayes Model, can help classify data using conditional probability models. Decision Trees are powerful classifiers and use tree splitting logic until pure or somewhat pure leaf node classes are attained. Random Forests apply Ensemble Learning to Decision Trees for more accurate classification predictions. Conclusion This completes ‘Classification’ tutorial. In the next tutorial, we will learn 'Unsupervised Learning with Clustering.'

In this project, a RFM model is implemented to relate to customers in each segment. Assessed the Data Quality, performed EDA using Python and created Dashboard using Tableau.

This code performs customer segmentation using RFM (Recency, Frequency, Monetary) analysis. It creates an RFM dataframe, determines optimal clusters with k-means, and orders clusters from worst to best. Customers are segmented into high, mid, and low value groups and visualized in a 3D plot for better marketing strategies.

Create advanced customer segments to drive better purchasing predictions based on behaviors. Using sales data, campaigns and promotions systems, this solution helps derive a number of features that capture the behavior of various households. Build useful customer clusters to target with different promos and offers.

The aim of this project is to create customer personas according to features and define segments according to a determined rule.

No description available

Using data: Customer's invoice file. Introductions: Customer Lifetime Value(CLTV) "Customer Lifetime Value is a monetary value that represents the amount of revenue or profit a customer will give the company over the period of the relationship". CLTV demonstrates the implications of acquiring long-term customers compare to short-term customers. Customer lifetime value (CLV) can help you to answers the most important questions about sales to every company: How to Identify the most profitable customers? How can a company offer the best product and make the most money? How to segment profitable customers? How much budget need to spend to acquire customers? CLTV indicates the total revenue from the customer during the entire relationship. CLTV helps companies to focus on those potential customers who can bring in more revenue in the future. CLTV = ((Average Order Value x Purchase Frequency)/Churn Rate) x Profit margin. Please check the below step for how to calculate CLTV. Algorithm: Step1: Calculate CLTV. Calculate the average order value of customers: Average order value = Total money spent / total number of transactions Calculate Purchase Frequency: Purchase Frequency = Total Number of Orders / Total Number of Customers Calculate Repeat rate and Churn rate: Repeat rate = How many customers have numbers of transactions more than one / total numbers of customers Churn rate = 1 - repeat rate Calculate the profit margin: Profit margin is the commonly used profitability ratio. It represents how much percentage of total sales has earned as the gain. Let's assume our business has approx 5% profit on the total sale. Profit margin = Total money spent on each customer * 0.05 Calculate customer lifetime value: Customer value = (Average Order Value * Purchase Frequency) / Churn rate Customer lifetime value = Customer value * Profit margin Step2: Predictive modelling. Build a regression model for existing customers. Take recent six-month data as independent variables and total revenue over existing time( here taking 2 years) as a dependent variable and build a regression model on this data. Pros and Cons of CLTV: CLTV helps you to design an effective business plan and also provide a chance to scale your business. CLTV draw meaningful customer segments these segment can help you to identify the needs of the different-different segment. Customer Lifetime Value is a tool, not a strategy. CLTV can figure out the most profitable customers, but how you are going to make a profit from them, it depends on your strategy. Generally, CLTV models are confused and misused. Obsession with CLTV may create blinders. Companies only focus on finding the best customer group and focusing on them and repeat the business, but it’s also important to give attention to other customers.

Unsupervised learning for market segmentation

We are going to apply machine learning on unlabeled retail data to create at least 5 customer segments.

# Bank Marketing Dataset ## Marketing Introduction: The process by which companies create value for customers and build strong customer relationships in order to capture value from customers in return. - Kotler and Armstrong (2010). Marketing campaigns are characterized by focusing on the customer needs and their overall satisfaction. Nevertheless, there are different variables that determine whether a marketing campaign will be successful or not. There are certain variables that we need to take into consideration when making a marketing campaign. ## The 4 Ps: 1) Segment of the Population: To which segment of the population is the marketing campaign going to address and why? This aspect of the marketing campaign is extremely important since it will tell to which part of the population should most likely receive the message of the marketing campaign. 2) Distribution channel to reach the customer's place: Implementing the most effective strategy in order to get the most out of this marketing campaign. What segment of the population should we address? Which instrument should we use to get our message out? (Ex: Telephones, Radio, TV, Social Media Etc.) 3) Price: What is the best price to offer to potential clients? (In the case of the bank's marketing campaign this is not necessary since the main interest for the bank is for potential clients to open depost accounts in order to make the operative activities of the bank to keep on running.) 4) Promotional Strategy: This is the way the strategy is going to be implemented and how are potential clients going to be address. This should be the last part of the marketing campaign analysis since there has to be an indepth analysis of previous campaigns (If possible) in order to learn from previous mistakes and to determine how to make the marketing campaign much more effective. ## What is a Term Deposit? A Term deposit is a deposit that a bank or a financial institurion offers with a fixed rate (often better than just opening deposit account) in which your money will be returned back at a specific maturity time. For more information with regards to Term Deposits please click on this link from Investopedia: https://www.investopedia.com/terms/t/termdeposit.asp ## Outline: 1. Import data from dataset and perform initial high-level analysis: look at the number of rows, look at the missing values, look at dataset columns and their values respective to the campaign outcome. 2. Clean the data: remove irrelevant columns, deal with missing and incorrect values, turn categorical columns into dummy variables. 3. Use machine learning techniques to predict the marketing campaign outcome and to find out factors, which affect the success of the campaign. ## Dataset Link https://archive.ics.uci.edu/ml/datasets/Bank+Marketing ## Dataset Information The data is related with direct marketing campaigns of a Portuguese banking institution. The marketing campaigns were based on phone calls. Often, more than one contact to the same client was required, in order to access if the product (bank term deposit) would be ('yes') or not ('no') subscribed. There are four datasets: 1) bank-additional-full.csv with all examples (41188) and 20 inputs, ordered by date (from May 2008 to November 2010), very close to the data analyzed in [Moro et al., 2014] 2) bank-additional.csv with 10% of the examples (4119), randomly selected from 1), and 20 inputs. 3) bank-full.csv with all examples and 17 inputs, ordered by date (older version of this dataset with less inputs). 4) bank.csv with 10% of the examples and 17 inputs, randomly selected from 3 (older version of this dataset with less inputs). The smallest datasets are provided to test more computationally demanding machine learning algorithms (e.g., SVM). The classification goal is to predict if the client will subscribe (yes/no) a term deposit (variable y). ## Attribute Information Input variables: #### bank client data: 1-age (numeric) 2-job : type of job (categorical: 'admin.','blue-collar','entrepreneur','housemaid','management','retired','self-employed','services','student','technician','unemployed','unknown') 3-marital : marital status (categorical: 'divorced','married','single','unknown'; note: 'divorced' means divorced or widowed) 4-education(categorical:'basic.4y','basic.6y','basic.9y','high.school','illiterate','professional.course','university.degree','unknown') 5-default: has credit in default? (categorical: 'no','yes','unknown') 6-housing: has housing loan? (categorical: 'no','yes','unknown') 7-loan: has personal loan? (categorical: 'no','yes','unknown') #### related with the last contact of the current campaign: 8-contact: contact communication type (categorical: 'cellular','telephone') 9-month: last contact month of year (categorical: 'jan', 'feb', 'mar', ..., 'nov', 'dec') 10-day_of_week: last contact day of the week (categorical: 'mon','tue','wed','thu','fri') 11-duration: last contact duration, in seconds (numeric). Important note: this attribute highly affects the output target (e.g., if duration=0 then y='no'). Yet, the duration is not known before a call is performed. Also, after the end of the call y is obviously known. Thus, this input should only be included for benchmark purposes and should be discarded if the intention is to have a realistic predictive model. #### other attributes: 12-campaign: number of contacts performed during this campaign and for this client (numeric, includes last contact) 13-pdays: number of days that passed by after the client was last contacted from a previous campaign (numeric; 999 means client was not previously contacted) 14-previous: number of contacts performed before this campaign and for this client (numeric) 15-poutcome: outcome of the previous marketing campaign (categorical: 'failure','nonexistent','success') #### social and economic context attributes 16-emp.var.rate: employment variation rate - quarterly indicator (numeric) 17-cons.price.idx: consumer price index - monthly indicator (numeric) 18-cons.conf.idx: consumer confidence index - monthly indicator (numeric) 19-euribor3m: euribor 3 month rate - daily indicator (numeric) 20-nr.employed: number of employees - quarterly indicator (numeric) Output variable (desired target): 21-y - has the client subscribed a term deposit? (binary: 'yes','no') ## License This dataset is public available for research. Citations - 1.Moro et al., 2014] S. Moro, P. Cortez and P. Rita. A Data-Driven Approach to Predict the Success of Bank Telemarketing. Decision Support Systems, Elsevier, 62:22-31, June 2014 2.Dua, D. and Graff, C. (2019). UCI Machine Learning Repository [http://archive.ics.uci.edu/ml]. Irvine, CA: University of California, School of Information and Computer Science.

Creating Customer Segments - 4th project for Udacity's Machine Learning Nanodegree

RFM is a method used for analyzing customer value, creating customer segments. It is commonly used in database marketing and direct marketing and has received particular attention in retail and professional services industries.

9 Email Marketing Tips For Content Marketers Even “agnostics” regarding email marketing can't hash out the following evidence - the average ROI from this promotional practice is close to 3,800%. Measureless opportunities to scale up and relative cheapness, compared to other reaching-out channels, are the two reasons why the email marketing is fair-haired by businesses. However, this is not about the price and physical extent alone. The chief advantage is a better alignment of communication with customers. If you hope a certain content strategy brings desirable results, overlooking the quality of mailing messages will be a sorry pitfall. Always keep in mind that newsletters, welcome, retention, and other emails are not just a brand's facade - but a powerful tool for generating conversions. By joining sides of email and content strategies, you can come up with synergy from both. In this guide, we’ll cover a few recommendations for content marketers on how to write email messages that work. Tips for email marketing Segment your list Split the batch of email recipients into smaller groups based on chosen criteria, and mail distinct relevant messages - for each. You can use recipients' GEO, demographic characteristics, or purchase history to distinguish homogeneous clusters and proceed with the content planning. Segmentation is the basic premise for personalization, and if you still doubt why bothering about the latter - here are just a few numbers we took from Instapage: 52% of customers claim they do care if the message was tailor-made or not 82% of marketers say that mail personalization increases the open ratio custom emails have 41% more unique clicks than mass-produced ones. To avoid a fragmented approach, use data from CRMs, website analytics tools, and other sources to define segments. Concerning phrasings, a good idea is to create Buyer personas profiles. Thus, you'll be able to choose the appropriate message length and wording. Say, design a newsletter to promote paid subscription for an email validator service. You've decided to distinguish corporate clients based on their company size and determined the following groups: #1 - B2Bs and #2 - sole entrepreneurs. Possible messages for the two: #1. Our "XXL" plan is perfect for agencies and enterprises. One can add unlimited users and conduct up to 100,000 checks per month. #2. With our "S" you get 1,000 credits and 5,000 unique recipients - for only $33 per month. Plus - a 7-days free trial. Use interactive content The best content marketers know that interactive content came into vogue a long time ago. As to emails, here are the most common examples: CSS animated buttons If you include CTAs buttons (that we hope you do) - liven them up a bit. Add an animated hover effect, so that every time a recipient puts a cursor on a button, it changes shape, shade, color, or text. “Add hover to emphasise objects”, source This shouldn’t necessarily be something dramatic - add tiny accents that will yet grab the user's attention. starring “Add a star rating component to engage readers with content”, source Including ranking or reviewing widgets in the email body is one of the most working ways to engage the reader with the message. Ask recipients to assess your product or service with stars. Add the link to Google Forms if you want to receive an extended opinion on overall customer satisfaction. pictures' rollovers “Use animated images to describe goods better”, source The effect is eagerly used by the ones who promote online stores. Using The rollover allows to show goods from different angles or even play with recipients, if relevant. Take into account that this feature only works on desktops - mobile mail users will see the very first picture only. images carousel “Add pieces of text directly on images”, source If you want to enhance goods cards with descriptive content, say - price and shipping details, use a carousel instead of a rollover. As so, you can add more info pictures to the email body and, hopefully, convert more recipients into customers. a countdown “Countdowns work well for limited in time offers”, source Again, this type of interactive content fits the online shopping niche. Animated clocks amplify urgency and theoretically increase conversions. But it's important to stay extremely careful and not to sound desperate - otherwise, the newsletter will end up in the recipient's "Spam". Improve design The attractiveness of an email is something granted on certain terms, indeed. Not all emails need to be flashy or include expensive designs. However, there are some prevailing common trends in the matter. By following them, you seem to show the recipient that your company is moving in step with the times, and not stuck in the 2000s. Here's the shortlist from the TOP email design trends list that a 99designs provides - as of 2021: magazine-styled “Make newsletters to look a bit editorial”, source More and more newsletters tend to look like a centerfold from good old printed media. With a strict following to the "Less is more" principle - clear fonts, short phrases, HD-quality images with a few objects on them, and short CTAs. hand-made illustrations “Unique pictures create a distinct flavour of your brand”, source Tailored icons or sketchy images - whatever fits your mailing purpose, just make sure it's not too bright, contrast, or overloaded with details. Give preference to clean colors. skeuomorphic objects This is when a design resembles a real object. To see an example - just open a reader App on your smartphone. “A skeuomorphic bookshelf”, source HD photographies “If you operate in the luxury segment, do not skimp on email visuals”, source These are expensive content, but if you work in fashion or other chick industries - it may be worth the effort. animated content Yeap, we've covered this in a previous tip. single scroll “Looks especially good on smartphones”, source Place the entire email content, including buttons, on the endless-looking long frame. Focus on conversions Stay focused on what's your mailing purpose. Don't forget that everybody expects to see a good ROI from email actions at the end of the reporting period. Craft effective CTAs - perceive these not as a sole button with a "Download now" text or so, but as an entire sense of a message that you write. To create a captivating CTA copy, adhere to the below advices: include win-win propositions Even though you’re not providing a customer with a discount or cash refund at the moment, your proposition may include a non-monetary incentive. New arrivals, selection of the latest news, free copies, advice from experts - the only rule here is to offer what’ll hold in high esteem. trigger on emotions Don't long-windedly list benefits. Instead, simulate a life situation and show how your product or service can help. use several CTAs throughout the email Email body may be viewed in several scrolls, especially when via small mobile devices’ screens. If you add a call to action at the beginning of the message, a mere number of users will get back to it after finishing reading the content. Thus, you may lose potential conversion. Include several buttons throughout the email body, but don’t sound repeatedly - change calls’ forms and wording. Encourage readers to reply Driving recipients to reply is challenging yet able to be done. First, choose the proper writing tone. According to an extensive study of emails that didn’t get a response, the most preferable is a 3rd-grade reading level. “Too elementary or too proficient tone may scare away readers”, source Of course, you must apply this recommendation with an eye on the recipient. If you mail to a professor or a government agency, a “3rd-grade” rule isn’t applicable. But all else being equal - simplify the lexicon to the level a schoolchild can understand it. Another trick is to sound overall happy. Emails that are enhanced with positive emotions get 10-15% more replies, on average than neutral ones. The best manner is to choose a slightly warm tone. Exaggerated excitement may look weird and even suspicious, especially when reaching out to business partners. And don’t forget about courtesy. A rare person will respond if you address him or her with a hair-raising “To whom it may concern” phrase. Make it personal Personification shouldn’t be confused with personalization. The second is rather about mailing fitting content from a commercial perspective, while the first term - about addressing the recipient as a one-off personality. Personal emails start with the recipient’s name - and no other way. They include references to the user's interests or past actions. For example, if your tourist agency’s client is interested in island vacations - you shall approach him or her with respective offers. They also shall contain personalized promotions, if any. The best way to expand this approach on hundreds or thousands of recipients is to launch trigger-based email campaigns. Create delivery scenarios for different segments or stages of a sales pipeline. Then prepare a fitting sequence of relevant content - for every single scenario. To give a human face to mailing, one can practice greetings, as well. Birthdays, state holidays, anniversaries, a new status in the loyalty system - there are a lot of examples of what one may congratulate the customer with. Keep your emails out of spam folders It is better not to launch mailing at all than to use an untrustworthy emails’ database. The risks are much higher than a slew of undelivered messages - from harming a sender's reputation to being banned by mailing systems. So it's better to stay proactive: tidy away broken, misspelled, temporary, or other worrisome emails from the database - either manually or with the help of software collect a valid email address only - through email finders avoid spam-trigger words establish a double opt-in validation set the correct mailing frequency. Make sure your emails look clean and crisp Newsletters shall afterall bring revenues - whether you want it or not. But in a bid of quantity, don’t lose the overall content integrity and sense: a subject line, pre-header, header, email body, and calls shall be consistent with one another the copy must be of the proper size; although the length depends on many factors, stick to an “ideal” interval - 50 to 125 words if can, don’t attach too many files or links to external websites - mailing filters are suspicious to these adapt the layout to fit smaller screens - nothing looks worse than broken email elements when you open it on mobile. Wrapping up It doesn't make much difference whether you create mailing content for personal or business purposes - these email marketing tips will serve both. No strains here - the recipient’s interest should be at your forefront. If you can hook him or her with the content by using tricks we've covered, you’ll never fail with enough conversions.

Problem Statement Introduction So far, in this course, you have learned about the Hadoop Framework, RDBMS design, and Hive Querying. You have understood how to work with an EMR cluster and write optimised queries on Hive. This assignment aims at testing your skills in Hive, and Hadoop concepts learned throughout this course. Similar to Big Data Analysts, you will be required to extract the data, load them into Hive tables, and gather insights from the dataset. Problem Statement With online sales gaining popularity, tech companies are exploring ways to improve their sales by analysing customer behaviour and gaining insights about product trends. Furthermore, the websites make it easier for customers to find the products they require without much scavenging. Needless to say, the role of big data analysts is among the most sought-after job profiles of this decade. Therefore, as part of this assignment, we will be challenging you, as a big data analyst, to extract data and gather insights from a real-life data set of an e-commerce company. In the next video, you will learn the various stages in collecting and processing the e-commerce website data. Play Video2079378 One of the most popular use cases of Big Data is in eCommerce companies such as Amazon or Flipkart. So before we get into the details of the dataset, let us understand how eCommerce companies make use of these concepts to give customers product recommendations. This is done by tracking your clicks on their website and searching for patterns within them. This kind of data is called a clickstream data. Let us understand how it works in detail. The clickstream data contains all the logs as to how you navigated through the website. It also contains other details such as time spent on every page, etc. From this, they make use of data ingesting frameworks such as Apache Kafka or AWS Kinesis in order to store it in frameworks such as Hadoop. From there, machine learning engineers or business analysts use this data to derive valuable insights. In the next video, Kautuk will give you a brief idea on the data that is used in this case study and the kind of analysis you can perform with the same. Play Video2079378 For this assignment, you will be working with a public clickstream dataset of a cosmetics store. Using this dataset, your job is to extract valuable insights which generally data engineers come up within an e-retail company. So now, let us understand the dataset in detail in the next video. Play Video2079378 You will find the data in the link given below. https://e-commerce-events-ml.s3.amazonaws.com/2019-Oct.csv https://e-commerce-events-ml.s3.amazonaws.com/2019-Nov.csv You can find the description of the attributes in the dataset given below. In the next video, you will learn about the various implementation stages involved in this case study. Attribute Description Download Play Video2079378 The implementation phase can be divided into the following parts: Copying the data set into the HDFS: Launch an EMR cluster that utilizes the Hive services, and Move the data from the S3 bucket into the HDFS Creating the database and launching Hive queries on your EMR cluster: Create the structure of your database, Use optimized techniques to run your queries as efficiently as possible Show the improvement of the performance after using optimization on any single query. Run Hive queries to answer the questions given below. Cleaning up Drop your database, and Terminate your cluster You are required to provide answers to the questions given below. Find the total revenue generated due to purchases made in October. Write a query to yield the total sum of purchases per month in a single output. Write a query to find the change in revenue generated due to purchases from October to November. Find distinct categories of products. Categories with null category code can be ignored. Find the total number of products available under each category. Which brand had the maximum sales in October and November combined? Which brands increased their sales from October to November? Your company wants to reward the top 10 users of its website with a Golden Customer plan. Write a query to generate a list of top 10 users who spend the most. Note: To write your queries, please make necessary optimizations, such as selecting the appropriate table format and using partitioned/bucketed tables. You will be awarded marks for enhancing the performance of your queries. Each question should have one query only. Use a 2-node EMR cluster with both the master and core nodes as M4.large. Make sure you terminate the cluster when you are done working with it. Since EMR can only be terminated and cannot be stopped, always have a copy of your queries in a text editor so that you can copy-paste them every time you launch a new cluster. Do not leave PuTTY idle for so long. Do some activity like pressing the space bar at regular intervals. If the terminal becomes inactive, you don't have to start a new cluster. You can reconnect to the master node by opening the puTTY terminal again, giving the host address and loading .ppk key file. For your information, if you are using emr-6.x release, certain queries might take a longer time, we would suggest you use emr-5.29.0 release for this case study. There are different options for storing the data in an EMR cluster. You can briefly explore them in this link. In your previous module on hive querying, you copied the data to the local file system, i.e., to the master node's file system and performed the queries. Since the size of the dataset is large here in this case study, it is a good practice to load the data into the HDFS and not into the local file system. You can revisit the segment on 'Working with HDFS' from the earlier module on 'Introduction to Big data and Cloud'. You may have to use CSVSerde with the default properties value for loading the dataset into a Hive table. You can refer to this link for more details on using CSVSerde. Also, you may want to skip the column names from getting inserted into the Hive table. You can refer to this link on how to skip the headers.

利用无监督学习(Kmeans, PCA等) 结合数据可视化来获取有效精准对应的客户分类,从而给以公司商业决策方面重要的参考依据.

Clustering of the customer activities for 24 hours by using K-means clustering feature in Tableau 10.Tableau 10 clustering feature automatically groups together similar data points. This type of clustering helps you create statistically-based segments which provide insight into how different groups are similar as well as how they are performing compared to each other.

Problem Description : To predict the Customer life time value for an auto insurance company based on different quantitative and qualitative features provided. Forecasting is an important approach to take an optimal decision and implement appropriate action plans. A major non-life insurance company wants to evaluate customer life time value based on each customer’s demographics and policy information including claim details. The CLV is a profitability metric in terms of a value placed by the company on each customer and can be conceived in two dimensions: the customer`s present Value and potential future Value. You are expected to create an analytical and modelling framework to predict the life time value of each customer based on the quantitative and qualitative features provided in the dataset and also cluster the train dataset to understand behaviour of each segment. Files : i. Train.csv (Build model) - ii. Test.csv ( Model will be tested on this in grader tool). iii. sample_submission.csv : Customer ID in the order in which the predictions need to be submitted and randomly generated sample predictions. Main Tasks : ● Data Pre-processing/Exploratory Data Analysis/Visualization on Train Data. ● Model Building ● Clustering and insights from clustering your data. This is left open-ended but you are expected to come up with your own ideas on which data subset is worth clustering and what insights you can get out of it. ● Presentation and Viva Primary Evaluation Metric : RMSE 50 mark(s)

Creating Customer Segments

With the coming of new innovation, it is very not entirely obvious out on regarded openings accessible. The present circumstance is surprisingly more dreadful when one doesn't have the aptitude to tap on these changes. Indeed, this is the situation for organizations which have restricted information on site advancement and plan. Let's be honest, site administrations have colossally changed how the business functions. Thus, for genuine business people or organizations wishing to know the significance of sites this article gives simply that. Makes route simple With regards to having an effective online stage, the client should appreciate simple route. Basically, data gave on the site ought to be anything but difficult to get to. Hence, it is normal that the pages have quick stacking speeds. Accordingly, the site like online car parts store is needed to offer alternatives to additional guide in route. This incorporates the consideration of an inquiry box. Here, the clients will type on the hunt instrument and rapidly be coordinated to the segment. It is through excellent website composition that a designer's site accomplishes this. Beside building up the site, the designer is encouraged to routinely test the pages for simplicity of route. This is to kill or resolve bugs that may hamper the simplicity of stacking site pages. Keep in mind, on the off chance that a site has great route abilities, at that point it is ensured of more natural traffic. Will win with SEO Site design improvement has become a major perspective to see with regards to the site. With a large number of sites challenging to top in web index results pages (SERPs), web crawlers needed to acquaint a path with list locales. All things considered, it is through web improvement and plan that one will achieve a higher positioning. Here, boundaries, for example, title labels, utilization of catchphrases, picture advancement, connecting among others are thought of. This suggests that the site fulfills all the guidelines needed by be positioned top. Thusly, it is through streamlining that the site becomes easy to understand. Beside having the site, the website admins will hold the truly necessary clients. Under this, the web engineer is needed to incorporate highlights, for example, "embolden". This further involves the need to have short sighted plans on the pages. In this way, you will learn on the normal stacking speeds. It is through this improvement that the site shows up when various questions are made. So the site gets more taps on query items. Give visual substance on the site Truth be stated, selling unique item and administrations can be unwieldy. This is additionally muddled when an organization just gives huge loads of text about their forte. It is here that site advancement flavours things up. By reaching an expert website specialist, the entrepreneur will pick the pictures to utilize. Furthermore, the undertaking has the opportunity to pick the quantity of promotion recordings and pictures. This will be guided by the improvement on web indexes. The value of utilizing visual substance is that furnishes the clients with an away from of what the item resembles. Evidently, not all clients comprehend the administrations or items offered through content. So the consideration of pictures makes it easy to drive the message home. Other than this, utilization of pictures on the site effectively catches the consideration of the peruses. Prior to perusing the content, clients are frequently enthused about the picture. This improves the odds of having more clients to the site. By the by, website admins are encouraged to try not to stuff the visual information. This is on the grounds that it makes it hard for the client to decipher. It additionally brings down the positioning of the site of site improvement. So it is imperative to direct the utilization of symbolism. Increment the deals Business flourishing is profoundly moored on the quantity of deals made. All things considered, making a site can adequately help an undertaking to draw in more deals. As per Statista, online business exercises are foreseen to develop by 21.3% constantly 2019. This demonstrates that deals on sites are drawing in more clients. These days, more entrepreneurs are hurrying to direct their exchanges on the web. This is on the grounds that they have detected the incredible chance to exploit online deals. The expansion in deals goes inseparably with the developing number of clients. To additionally advance the business, website admins are urged to incorporate updates. It is through updates and overhauls that the site capacities are smoothened. Also, it shows the customers that the brand is committed to offering excellent administrations and data. Another approach to improve the deals is by including advancements. Here, you will make the truly necessary fluff among clients. This repeats into more deals. Also, this gives clients the feeling that they can gain reasonable items from the organization. Along these lines, all exercises on the site increase the value of the business somehow. Draw in lifetime customers to your business As the organization tries to spread its wings and extend, it is crucial to have steadfast clients. In any case, this can be an overwhelming undertaking particularly when the business person utilizes helpless strategies to accomplish this. It is now that improvement and planning of the site help out. The measurements recovered from the webpage empower website admins to screen the action of clients. Here, it is conceivable to feature the clients that have persistently upheld the brand. In the wake of pinpointing them, the entrepreneur should utilize inventive approaches to hold these clients. One creative alternative is compensating them with blessing vouchers and prizes. This will give them more motivation to get to your administrations or items. Keep in mind, it is through the site that the entrepreneur guarantees no reliable client is forgotten about. Another interesting thing about the lifetime clients is that they can advertise the brand. So they get to in a roundabout way work for the organization. This additionally lessens the expense of showcasing. Contact more customers One of the principle objectives of building up an endeavor is to fill regarding client base. Indeed, there is a heap of approaches to accomplish this yet each has various outcomes. With regards to web advancement and plan, there are some significant achievements accomplished. The first is that it puts the brand name out there. Basically, when the site is free on Worldwide Web then the organization is on a worldwide stage. This implies that the generally secret endeavor can be looked and give items to far away clients. It is these administrations that guide to lessen the distance for the clients to get to the exercises. Here, there are different choices, for example, buying or requesting the item on the site. Besides, the organization actually stays in contact with the neighborhood clients. Incredible right! Improving client commitment Routinely, an undertaking was facilitated in a physical structure. Nonetheless, circumstances are different as more administrations have gotten advanced. It is thus that business visionaries are urged to create wonderful sites. In this stage, it is very simple to keep a decent affinity with the end client. This involves recovering input on the administrations and items advertised. So you can associate with them and give fundamental reactions to the inquiries inquired. Furthermore, there is no restriction on the hour of action. Via robotizing the administrations on the site, customers are ensured of nonstop administrations. Additionally under client commitment, the blog or webpage proprietor can update clients as often as possible consistently. For example, in the event that new value charges are presented, at that point clients are among the first to know. Creative in showcasing and promoting For new businesses, having items and administrations out there is major in making progress. All things considered, showcasing methodologies prove to be useful in selling the brand. Contrasted with strategies, for example, the utilization of primary media and boards, site advancement is pocket-accommodating. It is through this online stage that an organization can show all important data. This incorporates; items/administrations offered, area, estimating, notoriety, contacts among others. The website admin can helpfully post alluring proposals on the site. Strangely, it is simpler to refresh astounding limits and offers on the site. So there is no personal time in trusting that the promotion will be set up. A similar case applies when the organization wishes to pull down a blog entry or advert. Also, the undertaking can work with a given figure. I'm not catching this' meaning? Basically, through SEO the business can realize where to put more accentuation. Also, the site gives forward-thinking data on the most recent promotion on the lookout. Smoothing out the brand While presenting a site for the organization, it is critical that the brand name be predictable. It is through site advancement and website architecture that this is refined. Here, the website admin will make a particular brand name that will be included on all the web crawlers. So there is no variety whether or not the site is on Bing or Google. Moreover, the brand logo and name is comparable all through. This decreases the odds of disarray with other serious brands. This likewise streams down to the issue of consistency. It is foreseen that the organization keeps a consistent following of their clients. In the event of rebranding, the website admin ought to guarantee that the due system is followed. Whenever this is thought of, at that point the web indexes will naturally refresh the records. In this way, when clients look for the brand will get to the correct thing. One more approach to see this is that the site can help to advise customers regarding changes. As the organization utilizes different methods, for example, online media, the site can likewise contribute. Here, the website admin can even prod the peruses of another look before dispatch. All things considered, these progressions can be executed all through. How Website Development and Web Design Helps Enterprises to Make Profits Saving on expenses Shockingly, numerous start up and significant organization fall flat in their endeavour because of low benefits. This is regardless of having extraordinary assumption for the speculation made. Some portion of the disappointment is ascribed to commitment of helpless business strategies, for example, carelessness of web administrations. It ought to be realized that site advancement and configuration is reasonable. By appropriately organizing the substance, the website admin saves a great deal of cost during web advancement. The cost saving angle stretches out to the modern acquires the site will bring to the business. Moreover, the site lessens the distance covered to contact the clients. In the event that one was to actually converse with possible clients, at that point it would be asset concentrated. It is here that web administrations come in. Besides, reducing such additional expenses implies that the business is gathering more benefit. Allowing promotions on the site Entrepreneurs probably go over the numerous advertisements been communicated on different site. Indeed, this is one of the fascinating ways an undertaking can draw in more benefit. Fundamentally, the organization will be drawn closer by different ventures to have their advertisements run on the site. As a component of promoting and publicizing, the host site will charge a specific add up to have the advert. Accordingly, it is essential to think of an interesting and famous site. By zeroing in on this, the website admin will put the site on the spotlight. The huge champs here are those whose site pulls in more undertakings and promotions. E-Commerce As specified previously, pre-cuts and administrations have moved from the stores to online stages. One of the significant online scenes is the site. We should take the case of Amazon, figured out how to add to 44 percent of the complete web based business deals in the United States. In addition, Statista featured that the organization has figured out how to make $108.35 million out of 2017. After the top to bottom elaboration of the significance of a site, certain viewpoints come out clear. The first is that business undertakings should try to create and plan a custom site. Also, it is significant to put the best foot forward. So it is foreseen that the site or blog meets and outperforms the rules. Having said this, it is dependent upon the website admin to take that wide action and build up a site. Tag : website designing company in delhi, website development company in delhi

Applied Unsupervised Learning techniques on product spending data collected for customers of a wholesale distributor to identify customer segments hidden in the data.

This repo contains all my work for Project 2 of Udacity's Machine Learning Basic Nano-degree Program. In this project I applied unsupervised learning techniques on product spending data collected for customers of a wholesale distributor in Lisbon, Portugal to identify customer segments hidden in the data. I first explored the data by selecting a small subset to sample and determine if any product categories highly correlate with one another. Afterwards, I preprocessed the data by scaling each product category and then identifying (and removing) unwanted outliers. With the good, clean customer spending data, I applied PCA transformations to the data and implement clustering algorithms to segment the transformed customer data. Finally, I compared the segmentation found with an additional labelling and consider ways this information could assist the wholesale distributor with future service changes.

unsupervised learning techniques on product spending data collected for customers of a wholesale distributor

This project focuses on segmenting customers based on their tenure, creating "cohorts", allowing us to examine differences between customer cohort segments and determine the best tree based ML model.

We will segment customers based on their purchasing behavior patterns using a modified LRFM metric to create more more targeted and personalized promotions.

Creating a segment's of some customers from a bank based on their credit card details and divide them into some groups through clustering algorithms and then making a prediction on those groups for new customer's.

Introduction: India is entering in to E - Ages or 4G era of the Internet world is now developing very quickly. Internet has become a fundamental requirement all across the world in today’s business. With the help of Internet every small development in the industry is available on our fingertips through mobile and laptop computer. The importance of being active and live though internet increasing day by day. According to a survey There will be 259 millions of people using internet through mobile and laptop by end of this year. This leads to a conclusion that every common man is using the internet and connected himself through social media. Now, if we use this platform to promote our products through different social media like Facebook, Twitter, LinkedIn etc. then it will be a very fruitful marketing strategy at the end of the day to get the desirable results. About our company: OnlineTradeProfit (commodity and Jewelers Information Provider) is a service provider organization. We provide online marketing services through social Media for Commodity business, Jewelry business and all the businesses in this segment. List of services of OnlineTradeProfit: • Creating and developing the website for the business of any segment and industries. • Services related to Bulk SMSs and E-Mails • Services related to Domain and Hosting of a website • Services related to Social Media Marketing • Services related to marketing through bulk SMSs and e-mail for the huge customer data base of stock exchanged and their brokers Social Media Introduction: The recent trend of promoting and marketing of products through web sited like Facebook, Twitter, LinkedIn, and Google Plus and by creating blogs has increased a lot. We can market our products through properly developed Images or Graphics, Text, Audio or video. We can get the fruitful and desirable outputs by developing the company's database and it can be shown live through website. We also provide our expertise in this business if you already have developed your content and pages in the social media. Social media marketing contracts: To update your Facebook page, and multimedia design: (Which Google - Search engine optimizer and product marketing through image with graphics, text, audio or video) Scheme: Rs.10,000/- or Rs.15,000/- Monthly Sample: • https://www.facebook.com/investorgroupbse • https://www.facebook.com/vijayjewellers916 • https://www.facebook.com/IIJMAHD • https://www.facebook.com/PruthviDiamondCreation OLD Clients • https://www.facebook.com/SuvarnaShilpi • https://www.facebook.com/RamniklalandSons • https://www.facebook.com/GujaratInternationalJewelleryShow • https://www.facebook.com/GoyalJewelers • https://www.facebook.com/ChokshiMahajan • https://www.facebook.com/genAhead More Detail Call On 9824053541 Vipul Soni.

Tech career has taken over the world with an exponential increase in opportunities in the job market. Niche fields like data science and cybersecurity are booming with hefty salary packages. Every industry in the world is facing some technological disruption degree but, Information Technology is a diverse industry that can take careers in n number of directions. Which career path to follow in the world is not an easy determination? Here are the top 10 tech careers in Australia that are booming and taking off now, according to the Asia-Pacific Tech and Executive Talent Specialists Findings. Java, Python & GoLang Developers Their primary platforms work well with algorithms required to operate or build AI functionalities and Machine Learning. So they are in current top trends now. Experts in these are difficult to find by as the experience is mostly coming from the startup space, not from the larger employers. JavaScript Developers (with Angular or React Expertise) JavaScript Developers are among the top 10 tech careers in Australia. It’s a versatile language used across a variety of businesses so will continue to be in demand for their talents. The larger number of frameworks within the language makes it a talent pool segmented across niche expertise. DevOps (Cloud) & DevSecOps (Security) & Cloud Security Engineers These are required to maintain and automate the developer’s environment. It’s growing as businesses are becoming more reliant on high volumes of data. DevOps engineers speed up the deployment rate and ensure the minimal downtime of products. DevSecOps Engineers perform same with a focus on security code level security. Cloud Security Engineers instruct on maintaining the quality of coding methods which promote a secure cloud-based environment. Cyber Threat Intelligence / Cyber Emergency Response (CERT) Managers These specialists act as a bullet in the world of cyber attackers. They identify and mitigate risk and respond to cyber-attacks allowing businesses to maintain service, productivity, and data security. Automation Engineers They speed up the market and reallocate the resources by automating human capital savings and processes saving cost. Data Scientists These jobs are booming as businesses seek to better understand and derive value from the data they have generated from their customers. More enterprises are consequently turning to data scientists to apply data-driven strategy decisions and allocate resources more efficiently for a better forecast. Data Engineers They are the critical members for an enterprise as a data analytics team. Their demand is high as more and more businesses are becoming reliant on manipulation and collection of high volumes of data. Agile Coaches / Scrum Masters They are responsible for teaching and running methodological processes and frameworks in the search for efficiency. So demand for these roles will continue to grow. CX/UX Researchers and Designers They are responsible for collecting and making design decisions based on customer experience data and allow the businesses to create better efficiencies across the organization by optimizing their product to something that serves the requirements of the customers. Tech Sales Executives (Enterprise Level) Enterprise-level Tech Sales Executives will always be in demand as they are responsible for an efficient path towards the capital growth of the business. The salesperson who can build strong relationships and grow steadily are in constant demand.

Machine Learning Nanodegree Project Udacity

In this project I have used unsupervised learning techniques to see if any similarities exist between customers, and how to best segment customers into distinct categories using a real-world dataset

Applying Unsupervised Learning techniques to identify customer segments hidden in the data