머신러닝 알고리즘을 이용한 커피 프랜차이즈 실패 요인 분석에 관한 연구

The restaurant franchise headquarters and the franchisees are organically connected, so the failure of the franchisee affects the franchise headquarters. Their sales affect the unemployment rate, causing social and economic crises. This study aims to analyze factors that affect low sales stores that are closely related to franchise failure. To meet the objectives, we first collect Gangnam-gu POS(Point of Sales) data and then analyze them with various machine learning algorithms. Our results show that LGBM(Light Gradient Boosting Machine) has the highest performance (accuracy 0.908). We apply the results with SHAP(Shapley Additional exPlanations), which is an explainale AI, to visualize the positive and negative effects of variables. In the near future, this study is expected to be utilized in suggesting a store operation strategy that can reduce the probability of franchise closure.