Sales forecasting in the white goods industry: A data mining application

For companies with growth targets and continuous development to survive and survive; they need new insights, strategies and, most importantly, analysis of the data that can guide them. Thanks to these analyses, companies will be able to make the fastest and most accurate decision in an intensely competitive environment. In this study, the turnover and profitability status of the companies for the past years are analyzed and included in the data mining techniques; it is aimed to achieve the right result with artificial neural network, support vector machines and regression. By analyzing the data of previous years, it is aimed to have a foresground about the data of a new year. Accordingly, it is notible not to think that the level of profitability will increase as a result of the strategies that can be developed. The aim of this study is to prove this linear thinking mathematically and to move forward with solid steps towards the goal. In order to prove the situation, the company operating in the white goods sector; the referral data are taken on a monthly basis as dependent variable, real sector confidence index, gross domestic product, industrial production index, marriage rate, consumer confidence index and economic confidence index as independent variables. The data was tested in the Weka program during the implementation phase. Sales estimates were tested separately by three different methods and the results were compared with each other.