Application exploration of building deep learning model by adjusting algorithm combination relationship

According to the Convention, the basic data type and XML representation algorithm are adopted, and after the organization is completed, it is compiled into Python code.According to the vectors in mathematics, whether the given algorithms share the same type of variables, the forward propagation and back propagation iterative algorithms in deep learning are used to train the model, and a new algorithm is combined according to the value of vectors, which is similar to the forward calculation in deep learning.Using the sample data to run the algorithm, the vector code is adjusted according to the error value in order to achieve convergence, which is similar to the reverse learning process in deep learning.In this way, it is expected to train a suitable model only by adjusting the parameter combination relationship of the algorithm sequence, and the neuron nodes and weight parameters are greatly reduced, which is expected to bring better learning effect.