Credit-worthiness Prediction in Energy-Saving Finance Using Machine Learning Model

Abstract

Companies can form their own "ESCO model" with their capitals. Unfortunately, customer's creditworthiness was becoming more crucial for ESCO. Machine learning was used to predict the creditworthiness of clients in ESCO financing processes. This research aimed to develop a scoring model to leverage a machine learning and life cycle cost analysis (LCCA) to evaluate alternative financing for Energy Saving in Indonesia. The results of calculations using multinomial logistic regression showed that the accuracy value of prediction data with test data was 88.3562 %. The prediction rate result that refers to the percentage of correct predictions among all test data was 91.67%, and False Positive Rate (FPR) was 39.44%. The True Positive Rate was called Recall or 'Sensitivity Rate' as it was defined as several positive cases that were correctly identified (TPR) was 92.20%. We found the machine learning methods for creditworthiness prediction in retrofitting projects were fresh and worth a shot. It was hoped that this new practice would grow in popularity and become standard among ESCOs. Unfortunately, current machine-learning-based creditworthiness scoring practices lacked explain ability and interpretability. Unfortunately, ESCO must penalize the retrofitting project. As a result, since retrofitting was a new industry, the credit approval process was challenging to communicate to consumers. The most important thing for ESCO to deal with the project was to have a friendship and know-how with the client. Research from these case studies led to a clearer understanding of the factors affecting all parties' decisions to implement and continue with their ESCO project.