AI enterprise intelligent foreground prediction model expert knowledge base system based on intelligent financial big data

Abstract

The optimal implementation mode of the invention is (T11) '1, prediction data standard access, 2, prediction index system optimization, 3, prediction rule and logical reasoning construction, 4, prediction signal identification and evaluation, 5, prediction model construction, 6, business prediction scene combination, 7, linkage analysis result preview and audit, and 8, prediction result intelligent output'. The working principle and the working process of the AI enterprise intelligent foreground prediction model expert knowledge base system based on the intelligent financial big data are as follows: (T13) the working principle and the process of the AI enterprise intelligent foreground prediction model expert knowledge base system based on the intelligent financial big data comprise intelligent acquisition of prediction data, automatic calling of a prediction model, automatic selection of prediction parameters and intelligent output of a prediction result.

Description

technical field [0001] This invention / invention patent relates to a (T2) [according to the indicators, intelligently and automatically recommend the optimal model for prediction. The best forecasting methods intelligently automate batch forecasting tasks without human involvement. ], especially related to one (T3) [relationships among index forecasts, special forecasts, mixed forecasts and linkage analysis indices]. Background technique [0002] (T4) [Analytic model selection algorithm based on multiple business rules, accounting standards, formula standards, and multivariate equation calculation principles] discloses a [minimum extreme value of relative error, minimum extreme value of absolute error, maximum degree of fitting value, the smallest extreme value of fit, the largest F-score test value, and the smallest F-score test value], however, this / these designs have the following problems and shortcomings: (T5)【1. Diversity and accuracy of prediction method selection Di...

AI Technical Summary

Problems solved by technology

[0002] (T4) [Analytic model selection algorithm based on multiple business rules, accounting standards, formula standards, and multivariate equation calculation principles] discloses a [minimum extreme value of relative error, minimum extreme value of absolute error, maximum degree of fitting value, the smallest extreme value of fit, the largest F-score test value, and the smallest F-score test value], however, this / these designs have the following problems and shortcomings: (T5)【1. Diversity and accuracy of prediction method selection Difficult to choose

Images