An improved foa-bpnn poverty alleviation time prediction method

Abstract

The invention discloses an improved FOA-BPNN poverty-overcoming time predicting method. The method comprises the following steps: converting poverty-overcoming time prediction of a low income family into a mathematical problem, and constructing a BPNN predicting model; and constructing an improved FOA-BPNN time predicting model by combining with an improved fruit fly optimization algorithm with aBP neural network, so as to realize the prediction of the poverty-overcoming time of the low income family. According to the method, the population density and speed variable factors are introduced into a fruit fly algorithm, so that the capacity of jumping out a local optimal solution through the fruit fly algorithm is improved; the improved fruit fly optimization algorithm is combined with the BP neural network, a training error of the BP neural network is taken as an adaptation degree value of the fruit fly optimization algorithm, and an optimal initial parameter combination of the BP neural network is searched by virtue of the fruit fly optimization algorithm, so that the accuracy rate of the predicting model is increased.

Description

technical field [0001] The invention belongs to the technical field of big data application, and relates to an improved FOA-BPNN poverty alleviation time prediction method, in particular to a method for constructing a poverty alleviation time prediction model for poor households. Background technique [0002] As the current domestic research work focuses on the relevant theories of precise poverty alleviation from the perspective of sociology, and a large number of poverty alleviation and poverty alleviation methods have been proposed, but from the perspective of natural science, the deep-seated causes and mechanisms of poverty alleviation are accurately quantified. Analytical models and methods are rare. Therefore, driven by big data on poverty alleviation, the establishment of a standardized poverty alleviation prediction model with the help of popular methods such as deep learning has important theoretical significance and reference value for formulating more reasonable p...

AI Technical Summary

Problems solved by technology

[0004] At present, it is rare to use BP neural network and fruit fly optimization algorithm to construct a prediction model to realize the prediction of the time for poverty-stricken households to get out of poverty.

At the same time, the improvement of the fruit fly optimization algorithm focuses on the change of the search range caused by the iterative optimal value, ignoring the problem that the population density may affect the population diversity and thus limit the search accuracy.

Images