Automatic modeling operation system and operation method

Abstract

The invention discloses an automatic modeling operation system and an operation method. The system comprises a sample model selection unit which obtains problem type information corresponding to to-be-trained data according to the to-be-trained data, and selects a corresponding sample model according to the problem type information; a parameter range selection unit which selects corresponding historical hyper-parameters and parameter ranges corresponding to the historical hyper-parameters according to the sample model; a parameter adjusting unit which carries out training processing on the historical hyper-parameters and the parameter range based on the sample model to obtain trained hyper-parameters and an optimization model corresponding to the trained hyper-parameters; an evaluation unit used for carrying out evaluation on the optimization model, generating evaluation report information, realizing automatic modeling and model optimization, quantifying a parameter adjustment training process. High usability and high accuracy of the system are acquired. A high-accuracy training model result is acquired without manually establishing a model structure and a parameter range in a deep learning training process. And the process of establishing, optimizing and evaluating the model is simplified.

Description

technical field [0001] The present application relates to the technical field of deep learning, in particular, to an automatic modeling operation system and operation method. Background technique [0002] At present, the application scenarios of deep learning are becoming more and more extensive. Deep learning is heavily used in fields such as security, image recognition, prediction and estimation. In the process of using deep learning, we can divide the training process of deep learning into data preparation, model construction, parameter debugging and result evaluation. Data preparation is to collect and organize the training data and related test data needed to solve the problem; model construction is to select a deep learning model to solve the problem; parameter debugging is to screen a set of hyperparameters to make the selected model the best; result evaluation is to use some The test data evaluates the generated model and calculates how well the model performs. Amo...

AI Technical Summary

Problems solved by technology

[0005] The main purpose of this application is to provide an automatic modeling operation system and operation method to solve the problem that the traditional deep learning training process needs to manually establish the model structure and parameter range, and needs to rely on human experience to select model parameters based on each result evaluation Value adjustment, problems that cannot be quantified during debugging

Images