The Forecast Analysis for Household Accounting Farmers with the Neural Network

Abstract

The purposes of the research were to develop a model to forecast analysis for Farmer's household debt with the neural network.  The research was led by the Learning Theory of Neural Networks applied to the sample data, which was collected from the  farmers of Moo 5 Khok Sak Subdistrict, Bang Kaew District, Phatthalung from 1  May 2560  to 30 April 2561 to consists of 36,500 data sets, 17 factors, including, inMain, inSavings, inOther, cfood, cHouse, cElectric, cClothes, cVehicle, cMedicine, exlotterly, exGambling, exalcohol, excigaret,  debtcost,  cinteres,t ctravel, cmake andcdebt.

The research findings showed that the average accuracy is 99.1537%, the error of classification is 0.8463%. The results are very good and bring new knowledge from the research. The results can be applied to farmers of Thailand.