Research On Collaborative Stock Price Predicting With Intra-class Correlation

Stock market is an important part of economy, and a major choice of investment, so its stability is related to national economy and people’s livelihood. Therefore, research on stock price prediction algorithm has theoretical and practical significance.Current algorithms predict stock price with its own historical information, which exist model fitting error and artificial random fluctuations. On this basis, by modeling stock price and analyzing macro influence factors, this article puts forward the idea that predict price by multiple stocks, and constructs collaborative prediction algorithm and its architecture:stock classification by clustering algorithm, initial price prediction by current price prediction algorithm and initial price weighting by weighed algorithm; Then the article divides the algorithm into functional module and develops the collaborative prediction software, which can provide prediction results and process information. Finally, the article analyzes performance of the algorithm and its influence factors, which proves a better accuracy and stability.