Analysis of frequent trading effects of various machine learning models

In recent years, high-frequency trading has emerged as a crucial strategy in stock trading. This study aims to develop an advanced high-frequency trading algorithm and compare the performance of three different mathematical models: the combination of the cross-entropy loss function and the quasi-Newton algorithm, the FCNN model, and the vector machine. The proposed algorithm employs neural network predictions to generate trading signals and execute buy and sell operations based on specific condi...