INDEX FEATURE EXTRACTION-BASED STOCK INDEX PREDICTION METHOD, SERVER AND STORAGE MEDIUM

摘要

The present application provides a index feature extraction-based stock index prediction method and device and a storage medium. The method comprise: extracting index factors of all time points in a preset number of time series and corresponding yields; selecting, according to a preset rule, a number n of index factors to form an n-dimensional vector; and respectively forming the n-dimensional vector of each time point and the corresponding yield into sample data to be trained. Then, by using the n-dimensional vector in the sample data and the corresponding yield, performing training on the bidirectional long-term and short-term memory network model to determine model parameters. Finally, receiving a time series to be analyzed, and extracting the n-dimensional vectors of all time points of the time series and inputting the n-dimensional vectors into the trained bidirectional long-term and short-term memory network model to obtain a comprehensive explanatory index of the time series. With the present application, the features of the index can be deeply extracted, and the accuracy of stock index prediction can be improved.