A remanufacturing cost prediction model of used parts considering failure characteristics

•Developing a remanufacturing cost prediction model of used parts considering failure characteristics.•The model combines least squares support vector regression (LS-SVR) with semi-supervised learning method, which can make an accurate remanufacturing cost prediction with a small remanufacturing sample.•A used part's remanufacturing cost can be predicted before it has been remanufactured.•The model has an extensive value in application, and the machine tool and engineered machine industry are the likely user. Remanufacturing cost is directly concerned with the remanufacturability of used parts. However, due to the varieties of failure characteristics, remanufacturing costs are likely to vary greatly, even used parts with the same type. In order to predict the remanufacturing cost of used parts precisely, a novel prediction model based on least squares support vector regression and semi-supervised learning is proposed in terms of failure characteristics. In this model, a k-nearest neighbor (k-NN) algorithm is introduced to increase forecast precision by learning a small number of labeling samples and a large number of unlabeled samples. A case analysis shows that the prediction model has a good performance and is of great value in application and dissemination.