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).

A deep learning method for event driven stock market prediction. Deep learning is useful for event-driven stock price movement prediction by proposing a novel neural tensor network for learning event embedding, and using a deep convolutional neural network to model the combined influence of long-term events and short-term events on stock price movements

This repository hosts the code for the SAMBA model, proposed in our IEEE ICASSP paper "Mamba Meets Financial Markets: A Graph-Mamba Approach for Stock Price Prediction".

We implement a quantum-classical hybrid QLSTM model by incorporating quantum variational layers into the classical LSTM in order to improve the efficiency and trainability of LSTM for better stock price prediction.

Stock Price prediction using news data. The datasets used consists news and stock price data from 2008 to 2016. The polarity(Subjectivity, Objectivity, Positive, Negative, Neutral) data is gathered from the news data and further used to predict stock prices. Achieved an accuracy of 94% using XGBoost.

S&P 500 stock price prediction using Transformer model

The Stock Price Prediction App is a Streamlit-based web application that provides users with tools to analyze historical stock price data, visualize technical indicators, and make short-term price predictions using different machine learning models.

A collection of notebooks and different prediction models that can predict the stock prices. Also a comparison of how all these models performed.

Calculate technical indicators from historical stock data Create features and targets out of the historical stock data. Prepare features for linear models, xgboost models, and neural network models. Use linear models, decision trees, random forests, and neural networks to predict the future price of stocks in the US markets. Evaluate performance of the models in order to optimize them Get predictions with enough accuracy to make a stock trading strategy profitable.

An ANFIS Model for Stock Price Prediction

Stock price prediction model built using BERT and regression model trained on textual financial news data.

The project focused on the use of public data to assess the economic situation in the country based on the state of the stock market and national means of payment, in particular - of the national currency. As sources are used: Open data Ministry of Finance of the Russian Federation These Moscow Exchange Google Finance Data Technologies used: Backend: Databases (relational) - Microsoft SQL Server 2014 Databases (multivariate) models DataMining, OLAP-cube - Microsoft Analysis Services 12.0 Веб-сервер - Windows Server 2012 / Internet Information Services Самописный ASP.NET HTTP Restful интерфейс для взаимодействия с Frontend ETL (загрузка и пре-процессинг данных, управление обновлением данных) SQL Server Integration Services 2014 (разработка в Visual Studio 2013, SSDT) Frontend: AngularJS ChartJS Twitter Bootstrap These were chosen so that the detail (granularity) in the set is not less than 1 day. The result has been created and filled with data analytic repository (Kimball model, topology - star), which was used to build a multi-dimensional databases and OLAP-based cubes on it, as well as models of analysis of data on two main algorithms: Microsoft Time Series, Microsoft Neural Network . To ensure interoperability frontend and backend server for backend-server was set up HTTP-Restful interface JSON-issuing documents in the form of finished sets. The project includes two main areas: Intelligent visualization of open data Analysis of open data and the construction of forecasts based on them Intelligent visualization involves the use of MDX-queries to the OLAP-cube, followed by depression (drilldown) in the data, the system allows the user to quickly find the "weak points" of the economy, as part of the data collected. To predict the time a standard mix of algorithms ARTXP / ARIMA, without the use of queries involving cross-prediction (but it is possible to enroll in the system correct data). These algorithms have been tested primarily on foreign exchange rates (US dollar) and the assets of banks included in the special list of Ministry of Finance. In addition, for assets shows the different customization options algorithms - a long-term, short-term and medium-term (balanced) plan. Assessing the impact of oil prices and foreign currency exchange rate for the total market capitalization was conducted on a sample of the data collected: companies with a total market capitalization of 100 to 500 million rubles, present in the market during 2013-2015 Analytical server builds the neural network receiving the input exchange rates, companies, the weighted average share price, total capitalization of the company and the price of oil to requests received models give the opportunity to evaluate the growth rate of \ fall (if at all) the company's capitalization at historical exchange rates and / or the cost of oil. Built a system can expand to include new indicators, which will significantly increase the accuracy of forecasting.

ARIMA & GARCH models for stock price prediction

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 Market Price Prediction: Used machine learning algorithms such as Linear Regression, Logistics Regression, Naive Bayes, K Nearest Neighbor, Support Vector Machine, Decision Tree, and Random Forest to identify which algorithm gives better results. Used Neural Networks such as Auto ARIMA, Prophet(Time-Series), and LSTM(Long Term-Short Memory) then compare make Inferences about the model.

Applying Reinforcement learning models for stock price predictions

A stock price prediction model based on ARMA and GARCH

(1) LSTM-RNN stock prices (historical closing precies of S&P500) prediction using keras with tensorflow. (2) Experiments APIs on the network's hyper-parameters are provided through './mmodel/experiment.py'. (3) a website is built using this prediction model as engine with Flask and MySQL.

Market Risk Management with Time Series Prediction of Stock Market Trends using ARMA, ARIMA, GARCH regression models and RNN for time series analysis and prediction of short-term tends in stock prices.

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

Stockastic is an ML-powered stock price prediction app built with Python and Streamlit. It utilizes machine learning models to forecast stock prices and help investors make data-driven decisions.

基于金融知识图谱和情感分析的股价预测模型

An intelligent agent utilizing Large Language Models (LLMs) for automated financial news retrieval and stock price prediction.

Model news data in short, medium and long term for stock price trend prediction

Price prediction for Crypto, Stock, and Index using Hybrid Graph Attention Network (GAT) and Long Short-Term Memory (LSTM) models on PyTorch.

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.

Customize loss function by taking account of directional loss to make the LSTM model more applicable given limited resources

Visualising stock price predictions for several Machine Learning models using DASH, Python

Compilation of technical analysis tools (EMA, Bollinger bands), fundamental analysis, machine learning models (LSTM, Random forest, ARIMA, GARCH, Markov Regime Switching), traditional stock prediction tools (Monte Carlo), sentiment analysis (NLP) as well as portfolio optimization, with purpose to provide a better understanding regarding possible future price movements of the analyzed stock and propose an optimal asset allocation.

Crypto & Stock* price prediction with regression models.