Developing Cluster-Based Adaptive Network Fuzzy Inference System Tuned by Particle Swarm Optimization to Forecast Annual Automotive Sales: A Case Study in Iran Market

Automotive Industry has an important place all around the world and sales forecasting process supports companies to meet their goals such as sales revenue increase, efficiency improvement, capacity planning and customer care. Traditional methods such as time series and econometrics have been applied by scientists during last decades. However, recently sales forecast problem by means of machine learning techniques are welcomed by data scientists because of increasing power of information technology in both hardware and software aspects. In this research, the hybridization of clustering method, Adaptive network Fuzzy Inference System (ANFIS) and Particle Swarm Optimization (PSO) are developed to forecast annual automotive sales in Iran automotive market. Furthermore, in regard to evaluate the developed model, Artificial Neural Network (ANN) and ARIMA are introduced and comparative analysis in three different scenarios are provided. The results illustrate that proposed method outperforms the rest of techniques and would be more applicable in forecasting problem especially in uns table macro- environments.