Bank Soundness Level Prediction: ANFIS vs Deep Learning

The systemic nature of the risk bankruptcy financial institutions has become an important issue in maintaining existence and stability domestic global finance. use statistics for prediction so far provides optimal benefits. However, this approach limitations, especially since model is built based on systematic relationships, linearity normality aspects are often weaknesses. This can be overcome very efficiently through linear non-linear patterns by artificial intelligence models. One most popular these techniques Artificial Neural Network (ANN). Many studies show that ANN fuzzy set theory more accurate, adaptive, strong predicting compared to statistical technique integrate with logic systems Adaptive-Network-Based Fuzzy Inference System (ANFIS). ANFIS adaptive network functionally equivalent inference advantages logic. features its acclimatization capability where membership function parameters adapt change learning procedure. Utilizing still limited Indonesia. Therefore, study aims construct a institution much operational quickly, effective as hybrid ANN. results showed used predict best MAPE 0.140335507.