Method and system for parameter optimization and feature tuning for machine learning

Abstract

The present invention proposes a method for parameter optimization and feature tuning for machine learning, comprising the following steps: randomly generating multiple parameter sets; performing EnKF-based iterative optimization on multiple parameter sets; performance evaluation of parameter sets, and obtain a set pool and a supplementary parameter set according to the evaluation results, wherein, the performance of the parameter set in the set pool is higher than that of the parameter set in the supplementary parameter set; for the parameter set in the set pool and supplementary The parameter set in the parameter set is again subjected to iterative optimization and performance evaluation based on EnKF to obtain the optimal parameter set. The method of the invention can improve the calculation result and calculation efficiency of processing parameter optimization, and has strong versatility. The invention also provides a system for parameter optimization and feature optimization of machine learning.

Description

technical field [0001] The present invention relates to the technical field of parameter optimization of machine learning, in particular to a method and system for parameter optimization and feature optimization of machine learning. Background technique [0002] For general machine learning algorithms, the performance of the model mainly depends on its parameter configuration. Models generated with different parameter combinations often have large performance differences. Parameter optimization is a stochastic optimization problem, and its randomness is mainly reflected in: the training data and test data used to generate the model contain limited samples, which cannot reflect the whole, and the parameter space is based on an unknown joint distribution function. The basic definition of the problem is as follows: Given a training data set X T , where X T Based on the unknown data distribution G, the goal of parameter optimization is to find a parameter combination θ of the...

AI Technical Summary

Problems solved by technology

[0008] The scaling factor of the feature is used as a class of parameters to complete the feature tuning, and the optimization together with the parameters of the machine learning algorithm will lead to a large number of parameters

However, there is still a lack of fast, accurate, general-purpose, and effective techniques for machine learning parameter optimization, especially for high-dimensional continuous parameter space optimization.

Images