A CNN-LSTM deep neural network with technical indicators and sentiment analysis for stock price forecastings

The utilization of social media platforms, such as Twitter, has experienced a notable surge. Consequently, the sentiments expressed by both influential figures and the general public have come to exert a significant influence on public opinion. This influence extends to various domains, including the stock market, where traders are swayed by the ideas and posts disseminated through these channels. Hence, the present study seeks to improving accuracy in stock price for castings with sentiment analysis and utilization of technical indicators, thus an CNN-LSTM framework is introduced to forecast stock market with sentiment analysis and technical indicators. Three deep learning models along with technical indicators and sentiment analysis are used to forecast stock prices. The impact of time windows of different sizes; 1, 7, 14, 21, 30, 45 and 60 was compared on the performance of the model, and the best time window was selected to forecast stock prices. 4 error-based evaluation criteria have calculated. The best model to forecast the stock price was the CNN-LSTM model. The MSE of the model was 0.0011.