Next-Generation Stock Market Prediction: Integrating CNN and LSTM Model

Stock market forecasting is one of the most challenging tasks in the financial industry due to the non-parametric nature of the stock market's time series, which is also complex, dynamic, chaotic, volatile, noisy, and otherwise non-linear. Fortunately, with the help of modern computing, intelligent models can assist expert analysts and investors in reducing investment risk. Preprocessing, feature selection, and training the model are the three stages that make up this suggested approach. When preprocessing data, the empirical wavelet transform decomposition algorithm is employed. When it comes to feature selection, one unsupervised and competitive learning algorithm that has proven effective is the self-organizing feature map. The model was trained using a CNN-LSTM. The proposed technique outperforms CNN and LSTM with an average accuracy of 90.56 percent.