Method and system for parameter optimization of machine learning algorithm

Abstract

The invention provides a method and system for parameter optimization of a machine learning algorithm. The method comprises the steps: (A), determining the machine learning algorithm for training a machine learning model; (B), providing a graphical interface for a user to set a parameter adjustment configuration item of the machine learning algorithm, wherein the parameter adjustment configurationitem is used for limiting how to generate multiple groups of candidate algorithm parameter values; (C), receiving an input operation of a user for setting the parameter adjustment configuration itemon the graphical interface, and obtaining the parameter adjustment configuration item set by the user according to the input operation; (D), generating multiple groups of candidate algorithm parametervalues based on the obtained parameter adjustment configuration item; (E), training a machine learning model corresponding to each group of candidate algorithm parameter values under each group of candidate algorithm parameter values according to the machine learning algorithm; (F), estimating the effect of the trained machine learning model corresponding to each group of candidate algorithm parameter values.

Description

technical field [0001] The present invention generally relates to the field of artificial intelligence, and more specifically, relates to a method and system for tuning algorithm parameters for machine learning algorithms. Background technique [0002] At this stage, the basic process of training a machine learning model mainly includes: [0003] 1. Import a dataset (for example, a data table) containing historical data records; [0004] 2. Complete feature engineering, wherein various features are obtained by performing various processing on the attribute information of the data records in the data set (for example, may include combined features), and the feature vectors formed by these features can be used as machine learning samples; [0005] 3. Training the model, wherein, according to the set machine learning algorithm (for example, logistic regression algorithm, decision tree algorithm, neural network algorithm, etc.), the model is learned based on the machine learnin...

AI Technical Summary

Problems solved by technology

[0008] Exemplary embodiments of the present invention provide a method and system for tuning algorithm parameters for machine learning algorithms, so as to solve the problem in the prior art that the machine learning system for training machine learning models cannot be conveniently used in machine learning systems. Algorithms for automatic tuning of algorithm parameters

Images