Multimodal big data machine automatic learning system based on neural and symbolic

Abstract

The invention belongs to the fields of artificial intelligence, machine learning and cognitive computing, and specifically relates to a multi-modal big data machine automatic learning system based on nerves and symbols, aiming to solve the problem that existing machine learning methods are difficult to obtain high-quality data from dynamic evolution data. The problem of true mechanistic models. The system of the present invention: the feature engineering automatic construction module includes a data acquisition unit, an event hypergraph network automatic construction unit, and a network structure automatic update unit; the mechanism model automatic construction module includes a domain task definition unit, a model game design search unit, and a search grid acceleration optimization unit; the hyperparameter optimization module includes a hyperparameter initial space construction unit, a hyperparameter adaptive selection strategy unit, an adaptive optimization reasoning unit, a hyperparameter automatic migration unit, and a model data processing module. The present invention obtains a high-fidelity mechanism model through a big data machine automatic learning method and iterative fidelity evaluation, thereby improving the accuracy of complex system behavior cognition prediction.

Description

technical field [0001] The invention belongs to the fields of artificial intelligence, machine learning and cognitive computing, and in particular relates to a multimodal big data machine automatic learning system based on nerves and symbols. Background technique [0002] Big data has become the core element of resource allocation and optimization in the fields of global industrial production, circulation, distribution, consumption activities, and economic operations. Exploring the method of cognitive big data is an important research direction in the field of artificial intelligence. In the final analysis, big data records the independent operation mechanism of complex systems in the real world, as well as the dependence, competition, and association between complex systems and the environment. This knowledge is of great research value for timely and accurate grasp of national economic development, optimization of industrial structure, and advancement of social science gove...

AI Technical Summary

Problems solved by technology

[0005] In order to solve the above-mentioned problems in the prior art, that is, in order to solve the problem that the existing machine learning methods cannot automatically fragment the dynamic evolution data to obtain high-fidelity mechanism models, the first aspect of the present invention proposes a neural and symbol-based Multimodal big data machine automatic learning system, the system includes: feature engineering automatic construction module, mechanism model automatic construction module, hyperparameter optimization module, model data processing module;

Images