Data Analysis of Amazon Product Based on LSTM and GPR

Data Analysis of Amazon Product Based on LSTM and GPR

In this paper, we propose a method that combines models such as GPR with PSO optimization to predict the time series data. We use LSTM and TOPSIS with entropy weight method modification to process vari-ous types of data from various aspects, taking into account both tabular and textual data, and to mine valuable contents from them. Based on shopping data, we analyze the historical situation and predict the future sales of products. So that we can recommend the most suitable products for customers. At the same time, for merchants, this paper provides directions for product optimization and improvement of advertising and marketing strategies.