In order to improve the potential returns of stock investors, the analysis of FTS data and the discovery of changing laws are a major challenge for stock market analysis technology. With the continuous in-depth research of artificial intelligence by scholars from various countries, it has been widely used in the field of FTS forecasting, effectively processing data information, and timely predicting the trend development of stocks. The main purpose of this paper is to conduct research on financial time series (FTS) forecasting based on the stochastic differential equation (SDE) model. In order to illustrate the wide applicability of the SDE model to the financial field, this paper makes an empirical analysis of the stock market and the exchange rate market. The evolution equation of the MEP algorithm is selected, and the parameters are optimized by the GA and PSO algorithms. Experiments show that the correlation between the sentiment feature index and the closing price, the highest price, the lowest price, the opening price and the previous closing price is all greater than 0.4, and the correlation is moderately related. However, the correlation coefficient with the rise and fall and the rise and fall is only about 0.05, indicating that there is only a very weak correlation or nothing. For transaction volume and transaction amount, the correlation coefficient is around 0.3, indicating a weak correlation.