This project explores the efficiency of various recurrent neural network (RNN) models, specifically Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) models, in predicting stock market prices. The experiment includes different LSTM architectures such as 1D-LSTM, 2D-LSTM, and attention-based LSTM (ARO-LSTM), along with traditional GRU models. The objective is to evaluate the performance of these models on stock market time series data and identify the most effective architecture for accurate price prediction.
In the domain of financial forecasting, the choice of model architecture plays a crucial role in predicting stock prices accurately. This project aims to compare the efficiency of different RNN models, focusing on LSTM and GRU variants. The models tested include 1D-LSTM, 2D-LSTM, ARO-LSTM (attention-based LSTM), and traditional GRU.
The dataset used for training and evaluation consists of historical stock market prices. Stock technical dataset is directly scraped from Yahoo finance using yfinance.