Method for predicting result of electronic competitive event based on SW platform and deep learning

Abstract

The invention belongs to the technical field of competition prediction, and particularly relates to a method for predicting an electronic competition result based on a SW platform and deep learning. Comprising the following steps: collecting and preprocessing data; crawling related data according to the characteristics of the project, and storing, cleaning and processing the crawled data; constructing and training a model; a suitable machine learning or deep learning model is selected for prediction, and repeated iterative training is performed after data is introduced; predicting and optimizing a result; continuously optimizing and adjusting the model according to the comparison of a prediction result and an actual competition result; generating predictive text content; and on the basis of a template rule artificially made by feature engineering, generating prediction content in a specific text form from a real-time prediction result. According to the method, the computing power of the SW server can be fully utilized, the method is independent of the redundant electricity competition client, and efficient, accurate and real-time competition result prediction is realized in combination with the characteristics of the electricity competition and data analysis.

Images