Abstract:
A stock price index is a tool that indicates the price level and trends of the stock market. Many factors affect the volatility of the stock price. Therefore, accurate analysis of stock trading signals is challenging. This research aims to develop a forecasting model for stock price time series data combined with corporate news to predict stock price trends and to compare the performance of the deep learning model with the traditional machine model. The dataset focuses on corporate news and financial stock data by selecting eight corporates gathered from stocks in the Thai Industry Group Index and Sector Index (SET). The experiment used traditional machine learning models, namely Support Vector Machine and Multilayer Perceptron, and Deep learning models: Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), Bidirectional Long Short-Term Memory (BiLSTM) and Bidirectional Gated Recurrent Unit (BiGRU) for classifying daily stock trading signals into three classes: buy, sell, and hold. The result showed that the deep learning model performance provides better accuracy (0.93), precision (0.93), recall (0.93), and F1-score (0.92) than traditional machine learning models. Moreover, the model is applied to analyze initial investment approaches to find investment returns.