Hybrid-LSTM-GARCHE-model-for-NASDAQ-Stock-Prediction-Forecasting-Algorithm

To create a data-web application deployed using the azure app service, which was made on Streamlit, the leading Pythonic data application service. On this website, we display candlestick plots of various stocks listed on the NASDAQ, according to the option of the user; and utilize the Garch based time forecasting algorithm done using Seasonal arima model and conduct a virtual future prediction for the given stock, so as to be able to conduct non-pairs algorithmic trading using time forecasting and Garch-based deep learning.

The main models which are implemented in this project are:

1. Generalised AutoRegressive Conditional Heteroskedasticity (GARCH) is a statistical model used in analysing time-series data where the variance error is believed to be serially autocorrelated. GARCH models assume that the variance of the error term follows an autoregressive moving average process.

2. SARIMA- SARIMA is Seasonal ARIMA, or simply put, ARIMA with a seasonal component. ARIMA (Autoregressive integrated moving average) is a statistical analysis model that uses time-series data to either better understand the data set or to predict future trends. It consists of 3 components – Autoregressivre, Integrated and Moving Average.

Please check out the video link for explaination and project implementation: https://youtu.be/j_uGdLlRQco