Comprehensive Prediction of Stock Prices Using Time Series, Statistical, Machine Learning, and Deep Learning Models

• DOI: 10.36227/techrxiv.23618478.v1
• Publication Date: 2023-07-10T07:20:24Z
• Authors: Jaydip Sen; Abhishek Kumar; Aji Thomas; Nishant Kumar Todi; Olive Olemmyan; Swapnil Tripathi; Vaibhav Arora
• Keywords: Exponential Smoothing; Stock (firearms); Econometric model; Predictive modelling

<p>Over the years, researchers have strived to develop reliable and accurate predictive models for stock price prediction. The literature suggests that well-designed refined can provide painstakingly precise estimates of future values. This project aims showcase a comprehensive set predicting prices, including time series, econometric, statistical, machine learning-based approaches. dataset includes ten industry leaders from different sectors NIFTY50, spanning January 2017 December 2022.</p> <p>The models' performance was evaluated determine which approach performs best sectors. series employed include Holt's Linear Trend Holt-Winters Exponential Smoothing, while econometric model utilized is ARIMA. Additionally, statistical adopted OLS, several learning deep incorporated range techniques such as Random Forest, KNN, CNN, LSTM, etc. Besides were also designed direction movements. insights into methods prices across industries by combining approaches.</p>