Jupyter Notebook tutorials on solving real-world problems with Machine Learning & Deep Learning using PyTorch. Topics: Face detection with Detectron 2, Time Series anomaly detection with LSTM Autoencoders, Object Detection with YOLO v5, Build your first Neural Network, Time Series forecasting for Coronavirus daily cases, Sentiment Analysis with BER

Project analyzes Amazon Stock data using Python. Feature Extraction is performed and ARIMA and Fourier series models are made. LSTM is used with multiple features to predict stock prices and then sentimental analysis is performed using news and reddit sentiments. GANs are used to predict stock data too where Amazon data is taken from an API as Generator and CNNs are used as discriminator.

time series forecasting using pytorch，including ANN,RNN,LSTM,GRU and TSR-RNN，experimental code

Stanford Project: Artificial Intelligence is changing virtually every aspect of our lives. Today’s algorithms accomplish tasks that until recently only expert humans could perform. As it relates to finance, this is an exciting time to adopt a disruptive technology that will transform how everyone invests for generations. Models that explain the returns of individual stocks generally use company and stock characteristics, e.g., the market prices of financial instruments and companies’ accounting data. These characteristics can also be used to predict expected stock returns out-of-sample. Most studies use simple linear models to form these predictions [1] or [2]. An increasing body of academic literature documents that more sophisticated tools from the Machine Learning (ML) and Deep Learning (DL) repertoire, which allow for nonlinear predictor interactions, can improve the stock return forecasts [3], [4] or [5]. The main goal of this project is to investigate whether modern DL techniques can be utilized to more efficiently predict the movements of the stock market. Specifically, we train a LSTM neural network with time series price-volume data and compare its out-of-sample return predictability with the performance of a simple logistic regression (our baseline model).

No description available

time series forecasting using keras, inlcuding LSTM,RNN,MLP,GRU,SVR and multi-lag training and forecasting method, ICONIP2017 paper.

Time series forecasting using LSTM in Python

Perform multivariate time series forecasting using LSTM networks and DeepLIFT for interpretation

This project forecasts renewable energy demand using LSTM-based time series models. It processes historical demand data, trains predictive models, and visualizes future trends, enabling better planning and management

Predicting future temperature using univariate and multivariate features using techniques like Moving window average and LSTM(single and multi step))

使用LSTM、GRU、BPNN进行时间序列预测。Using LSTM\GRU\BPNN for time series forecasting. (Pytorch Edition)

Using K-NN, SVM, Bayes, LSTM, and multi-variable LSTM models on time series forecasting

Financial Time Series Price forecast using Keras for Tensorflow. RNN LSTM

使用支持向量机、弹性网络、随机森林、LSTM、SARIMA等多种算法进行时间序列的回归预测，除此以外还采取了多种组合方法对以上算法输出的结果进行组合预测。Support vector machine, elastic network, random forest, LSTM, SARIMA and other algorithms are used for regression prediction of time series. In addition, a variety of combination methods are used to forecast the output of the above algorithms.

Using LSTM network for time series forecasting

No description available

This project uses an LSTM neural network to predict air quality (PM2.5) from synthetic time-series data. It includes data generation, normalization, model training, and prediction visualization. The results demonstrate how deep learning can forecast pollution levels

Time-Series forecasting using Stats models, LightGBM & LSTM

Time-series forecasting with 1D Conv model, RNN (LSTM) model and Transformer model. Comparison of long-term and short-term forecasts using synthetic timeseries. Sequence-to-sequence formulation.

Time-series demand forecasting is constructed by using LSTM, GRU, LSTM with seq2seq architecture, and prophet models.

Predict seasonal item sales using classical time-series forecasting methods like Seasonal ARIMA and Triple Exponential Smoothing and current methods such as Prophet, Long Short-Term Memory (LSTM), and Convolutional Neural Network (CNN)

This machine learning model (LSTM Time Series model) helps us to forecast demand of a supply chain business problem. This model uses Keras which uses tensorflow to solve the problem.

this project is to implement different deep learning architectures and evaluate them based on their performance on the hour-ahead electricity price and load prediction task. More specifically, we will evaluate (i) Random Forest, (ii) CNN-Univariate, (iii) CNN-Multivariate, (iv) RNN-LSTM and (v) BiLSTM architectures, using the root mean squared error (RMSE). Furthermore, we will experiment on different task formulations and types of frameworks, alongside the two following dimensions: • We will compare the performance of univariate time series forecasting and multivariate time series forecasting. Univariate time series forecasting is a framework on which the predicted quantity (i.e. electricity price) is the sole feature that is used by the models, whereas the multivariate variant of the task also uses other features which may prove important for the prediction, such as the load of the energy grid, the temperature, etc. • We will compare the performance of using different time-steps (3, 10 and 25 time-lags) as a way of reframing the time-series prediction task into a supervised learning problem, i.e. using the past 3, 10 and 25 values of the features which are fed into our models.

A time series forecasting project from Kaggle that uses Seq2Seq + LSTM technique to forecast the headcounts. Detailed explanation on how the special neural network structure works is provided.

This project implements a time series multivariate analysis using RNN/LSTM for stock price predictions. A deep RNN model was created and trained on five years of historical Google stock price data to forecast the stock performance over a two-month period.

Stock markets are an essential component of the economy. Their prediction naturally arouses afascination in the academic and financial world. Indeed, financial time series, due to their widerange application fields, have seen numerous studies being published for their prediction. Some ofthese studies aim to establish whether there is a strong and predictive link between macroeconomicindicators and stock market trends and thus predict market returns. Stock market prediction howeverremains a challenging task due to uncertain noise. To what extent can macroeconomic indicatorsbe strong predictors of stock price ? Can they be used for stock trends modeling ? To answer thesequestions, we will focus on several time series forecasting models. We will on the one hand usestatistical time series models, more specifically the most commonly used time series approachesfor stock prediction : the Autoregressive Integrated Moving Average (ARIMA), the GeneralizedAutoregressive Conditional Heteroscedasticity (GARCH) and the Vector Autoregressive (VAR)approach. On the other hand, we will be using two deep learning models : the Long-Short TermMemory (LSTM) and the Gated Recurrent Unit (GRU) for our prediction task. In the final section ofthis paper, we look directly at companies to detect trends

Time Series Forecasting of Bitcoin Prices using LSTM and RNN with Particle Swarm Optimization and Grey Wolf Optimizer

The research provides effective management strategies for different asset portfolios in the financial sector by building models. The VMD-LSTM-PSO model is developed for daily financial market price forecasting, where the time series are decomposed by VMD and the sub-series are used as LSTM input units to carry out forecasting, and then the network parameters are adjusted by PSO to improve the forecasting accuracy, and the Huber-loss of the model is 1.0481e-04. For the daily portfolio strategy, EEG is used to construct a system of investment risk indicators, which is optimized by incorporating the risk indicators into the Sharpe index, and the objective function is analyzed by using GA to derive the optimal daily asset share that maximizes the investor's return with minimal risk. The results of the empirical analysis show that the model provides strategies with good robustness.

Designing a Machine Learning algorithm to predict stock prices is a subject of interest for economists and machine learning practitioners. Financial modelling is a challenging task, not only from an analytical perspective but also from a psychological perspective. After 2008 financial crisis, many financial companies and investors shifted their interest towards predicting future trends. Most of the existing methods for stock price forecasting are modelled using non-linear methods and evaluated on specific data sets. These models are not able to generalize for diverse datasets. Financial time series data is highly dynamic in nature and makes it difficult to analyze through statistical methods. Recurrent Neural Networks (RNN) based Long Short- Term Memory (LSTM) networks were able to capture the patterns of the sequences data meanwhile statistical methods tried to generalize by memorizing data instead of recognizing patterns. In this work, we examined the performance of LSTM model and statistical models over stock prices of different companies to generalize the model. The experimental results of this study show that, LSTM network outperformed traditional statistical methods like ARIMA, MA and AR models. Furthermore, we have noticed that, LSTM network was able to perform consistently on different data sets while statistical methods showed varied performance. Through this project, we addressed the gaps in current models of stock price prediction in both economic and machine learning perspective.

In this project we have explored the use of imaging time series to enhance forecasting results with Neural Networks. The approach has revealed itself to be extremely promising as, both in combination with an LSTM architecture and without, it has out-performed the pure LSTM architecture by a solid margin within our test datasets.