مدل‌سازی پرداخت قبض آب در بخش خانگی شهر قم

Overdue water and wastewater bills of clients impedes any effort by utility companies at providing better services. It is thus essential to identify the factors affecting the timely payment of water bills. The data used in this study consists of the information of the first five billing periods of 2019 and 2020 for 181578 household clients in Qom, Iran. The logistic regression and decision tree models are used for the analysis of this data. The results showed that the methods have similar performance and the accuracy of the models is close to each other and equal (72%). The highest accuracy in late and timely payments is related to logistic regression (72%) and decision tree (75%), respectively. The two methods of support vector machine and neural network have similar performance. Logistic regression method indicates that billing payment status in the previous year and the total debt amount at the end of the previous year, as well as the number of unmetered bills and the average consumption of the current year, with regression coefficients of 1.75, -0.86, -0.70 and 0.034 respectively, are the most significant factors affecting the number of timely paid bills. The results of the decision tree model show that 43% of subscribers did not pay on time in the new year, did not pay on time in the previous year. Also, 39% of subscribers who pay on time in the new year, also paid on time in the previous year and in the new year they did not have unmetered bills. Comparing the results of logistic regression and decision tree models has shown that both methods have similar accuracies. The results show that the payment behavior in the previous year, the total amount of debt at the end of the previous year, and the number of unmetered bills in the current year have a significant effect on the number of paid bills in the current year. Therefore, various monitoring, incentive, and preventive measures can be used to improve the prospect of collecting the receivables. Increasing the supervision on meter reading staff, surveying the customers with unmetered bills, providing bonuses for timely payments, installment of previous debts, assigning penalties for delayed payments, and finally, cutting the water supply as a last resort are among these measures. In addition, the Omid utilities executive order, issued in January 2021, indicates that the household customers with monthly consumption of less than 5 m³ are exempt from payment. Therefore, these customers must be monitored, and special measures should be taken to collect their outstanding debts.