/Stock-Prediction-Portfolio-Optimization

A Streamlit based application to predict future Stock Price and pipeline to let anyone train their own multiple Machine Learning models on multiple stocks to generate Buy/Sell signals. This is a WIP and I will keep on adding new ideas to this in future.

Primary LanguageJupyter Notebook

Feature Extraction & Stock Price Prediction Pipeline (for training on Multiple Stocks)

A Streamlit based application to extract features and predict future Stock Price. In this application, I have developed a Pipeline to let anyone train their own multiple Machine Learning models on multiple stocks simultanously to generate Buy/Sell Signals using the best model.

There are 3 Dashboards in the web application -

1. Stock Exploration & Feature Extraction

2. Train ML Models for next day Stock Price Prediction

3. Train LSTM network for future Stock Price Prediction beyond just next day

This is a WIP and I will keep on adding new ideas to this in future.

Streamlit WebApp link - https://stock-prediction-dashboard.herokuapp.com/

Medium Post - https://medium.com/@nikhilkohli1992/extracting-features-for-stock-prediction-streamlit-based-application-a97afc55d926

Its a Numbers Game

Dashboard Screenshots -

Streamlit Dash

Streamlit Dash

Streamlit Dash