A Neighborhood Deep Neural Network Model using Sliding Window for Stock Price Prediction

Stock price prediction plays a crucial role in building a trading strategy for investors. The successful forecasting of stocks' future price will help the investors to increase their profit. However, it is difficult to predict exactly the trend of the stock market due to the complex relationship between stock prices and external factors such as news, global economy, public sentiments, and other sensitive financial information. Hence, historical prices are the main investigation data for researchers since collecting external factors is still a challenge. In this paper, we propose a novel method using the state-of-the-art deep neural network for time series prediction (long short-term memory - LSTM) along with deep concern in historical data of stocks and their nearest neighbors in terms of similarity. Experimental results on four active stocks in the United State and three stocks in the Vietnam stock market show that our model outperforms the competitive methods (vanilla LSTM, convolutional neural network - CNN, random forest, and linear regression).