TY  - JOUR
TI  - Sales History-based Demand Prediction using Generalized Linear Models
AB  - It’s vital for commercial enterprises to accurately predict demand by utilizingthe existing sales data. Such predictive analytics is a crucial part of their decision supportsystems to increase the profitability of the company.In predictive data analytics, the branchof regression modeling is used to predict a numerical response variable like sale amount. Inthis category, linear models are simple and easy to interpret yet they permit generalizationto very powerful and flexible families of models which are called Generalized linearmodels (GLM). The generalization potential over simple linear regression can be explainedtwofold: First, GLM relax the assumption of normally distributed error terms. Moreover,the relationship of the set of predictor variables and the response variable could berepresented by a set of link functions rather than the sole choice of the identity function.This work models the sales amount prediction problem through the use of GLM. Uniquecompany sales data are explored and the response variable, sale amount is fitted to theGamma distribution. Then, inverse link function, which is the canonical one in the caseof gamma-distributed response variable is used. The experimental results are comparedwith the other regression models and the classification algorithms. The model selection isperformed via the use of MSE and AIC metrics respectively. The results show that GLMis better than the linear regression. As for the classification algorithms, Random Forestand GLM are the top performers. Moreover, categorization on the predictor variablesimproves model fitting results significantly.
AU  - Tekir, Selma
AU  - ÖZENBOY, BAŞAR
DO  - 10.19113/sdufenbed.558620
PY  - 2019
JO  - Süleyman Demirel Üniversitesi Fen Bilimleri Enstitüsü Dergisi
VL  - 23
IS  - 3
SN  - 1300-7688
SP  - 840
EP  - 849
DB  - TRDizin
UR  - http://search/yayin/detay/346024
ER  -