Option Pricing Model Combining Ensemble Learning Methods and Network Learning Structure

Figure 1

Illustration of the deep ensemble model framework. The model consists of two parts, namely, features reorganization module and multilayer ensemble module. In the multilayer ensemble module, the output of each layer is spliced with the original input data as the input of the next layer; the specific number of layers is determined by the stopping algorithm , when , layers stop growing automatically, determined, and and represent two different ensemble learning methods, and represents historical data from day to day , i.e., .