AFTER-IQEA combination forecasting model for cosmetics sales forecasting

Cosmetics is necessary for everyone´s daily live, its impact on economy can not be ignored, but severe inventory stacking and lacking problems still exist. However, the occurrence of these problems is likely to be decreased via forecasting demand accurately. Thus, an Aggregated Forecast Through Exponential Re-weighting-Improved Quantum Evolutionary Algorithm (AFTER-IQEA) forecasting model is developed in this research. Important influential factors are chosen by Gray Relation Analysis, while considering the seasonal factor by Winter´s exponential smoothing. Three models: Evolving Neural Network (ENN), Adaptive Network-based Fuzzy Inference System (ANFIS), and Particle Swarm Optimization Wavelet v-Support Vector Machine (PSOWv-SVM) are used to forecast separately, and then integrate into AFTER-IQEA by dynamic weights generated from AFTER algorithm. The effectiveness of the proposed approach is demonstrated using real-world data, and it is superior to other traditional statistical models and neural network.