Data processing model and performance analysis of cognitive computing based on machine learning in Internet environment

摘要

To solve the problem that the rapid growth of data scale in the Internet environment makes it more difficult to extract meaningful information from massive data, a cognitive computing model is proposed, and the data processing technology that the model can use and the task scheduling algorithm under distributed background conditions are explored. In addition, aiming at the decision-making problem of the availability of massive information in the Internet, a cognitive decision algorithm based on deep confidence network and linear perceptron is proposed. According to the result of deep confidence network training, a cognitive decision model with error control function is established, and the decision result of whether the information is effective or not is given by considering the information itself and context information comprehensively. The experimental results show that the cognitive decision model based on deep belief network and linear perceptron performs well in the late training period. In addition, according to the experimental results, the most suitable operating parameters are selected. The accuracy of the algorithm is much higher than that of the deep belief network and the multilayer perceptron, indicating that this algorithm has some improvement in cognitive decision accuracy.