Transfer learning applied to bivariate forecasting on product warranty data

The reliability and resource management of products for warranty is important. Furthermore, the number failures aproduct over time use level expenditure can assume different distributions. Approaches with parametric modelsbring good results when there a normal distribution, application Deep Learning (DL) very promising. Weshow new methodology DL models transfer learning to bivariate forecasts repair rates inproducts that are under warranty. solution was applied data from an American company, recorded 2015 to2022, 12 types parts 69 cars. An evaluation absolute error wasperformed each combination part, car model year. Tests showed performed well in predictingdata 70 months service 70,000 miles, using cars at least 15 1,000 milesas input. It also concluded robust cases incomplete distributions far thenormal distribution.