Accurately predicting common stock trends plays an important role in making informed decisions in the stock market. With the ever-increasing complexity of the stock market, traditional statistical methods often fail to capture the dynamics of the market. Hence, the use of machine learning models has gained popularity in recent years. Among these models, Long short-term memory (LSTM) is a specific type of RNN model that has shown exceptional performance in handling sequential data. LSTM overcomes the limitations of traditional RNN models by solving the problem of gradient vanishing and exploding gradient problems.
This paper aims to explore the effectiveness of LSTM (using Keras Python) in predicting stock movement data. To this end, the study compares LSTM's performance with the use of CNN and GRU models in predicting the same data. The comparison helps to identify the model that performs best in predicting stock movements.
Furthermore, the study develops four different hybrid models to fine-tune the existing models. These models combine the strengths of each approach to maximize the prediction performance. The development of the hybrid models allows us to optimize the existing models to achieve better results than the individual models.
Overall, this study provides valuable insights into the effectiveness of LSTM models in predicting stock trends accurately. The findings of this study can help investors make informed decisions in the stock market, leading to better investment outcomes.
cliffkwok / ltsm_msft_paper Goto Github PK
View Code? Open in Web Editor NEWThis is my revised project for the finance with AI project