Developing a Novel Machine Learning-Based Classification Scheme for Predicting SPCs in Colorectal Cancer Survivors

Abstract

Colorectal cancer is ranked third and fourth in terms of mortality and cancer incidence in the world. While advances in treatment strategies have provided cancer patients with longer survival, potentially harmful second primary cancers can occur. Therefore, second primary colorectal cancer analysis is an important issue with regard to clinical management. In this study, a novel predictive scheme was developed for predicting the risk factors associated with second colorectal cancer in patients with colorectal cancer by integrating five machine learning classification techniques, including support vector machine, random forest, multivariate adaptive regression splines, extreme learning machine, and extreme gradient boosting. A total of 4287 patients in the datasets provided by three hospital tumor registries were used. Our empirical results revealed that this proposed predictive scheme provided promising classification results and the identification of important risk factors for predicting second colorectal cancer based on accuracy, sensitivity, specificity, and area under the curve metrics. Collectively, our clinical findings suggested that the most important risk factors were the combined stage, age at diagnosis, BMI, surgical margins of the primary site, tumor size, sex, regional lymph nodes positive, grade/differentiation, primary site, and drinking behavior. Accordingly, these risk factors should be monitored for the early detection of second primary tumors in order to improve treatment and intervention strategies.

Hybrid Basketball Game Outcome Prediction Model by Integrating Data Mining Methods for the National Basketball Association

“…MARS consists of a series of weighted sum of the basis functions (BFs), which are splines piecewise polynomial functions, and are demonstrated in the following equation [ 18 , 27 ]: where and are constant coefficients that can be estimated using a least-squares method. is the total number of BFs.…”

“…ELM, a single-hidden-layer feedforward neural network that randomly determines the input weights and systematically computes the output weights of the network [ 28 ], has a faster modeling time than the conventional feedforward network learning algorithms. It reduces usual disadvantages found in gradient-based methods, such as stopping criteria, learning rate and epochs [ 18 ].…”

“…Therefore, a linear system achieves its minimum norm least-square solution by , where represents the Moore–Penrose generalized inverse of the original matrix . The minimum norm least-squares solution has the smallest norm among all other solutions [ 18 , 29 ].…”

“…Only the authors of [ 13 ] implemented data mining methods on NBA game score prediction. The five well-known data mining methods, including multivariate adaptive regression splines (MARS), k-nearest neighbors (KNN), extreme learning machine (ELM), eXtreme gradient boosting (XGBoost), and stochastic gradient boosting (SGB), are used in this study for building an NBA game score prediction model, as they have been widely used in various applications such as public health [ 17 , 18 ], finance [ 19 , 20 ] and civil engineering [ 21 , 22 ]. Moreover, the five methods are also successfully applied in the sports outcomes prediction research [ 6 , 23 , 24 , 25 , 26 ].…”

Risk Prediction for Early Chronic Kidney Disease: Results from an Adult Health Examination Program of 19,270 Individuals

“…Early CKD awareness is essential for potential patients to participate in and comply with adult preventive health examination programs. Indeed, data mining has been successfully used for building a predictive model for healthcare prediction tasks [ 16 , 17 , 18 , 19 , 20 ]. Thus, in this study, four data mining algorithms, including a classification and regression tree (CART), a C4.5 decision tree, a linear discriminant analysis (LDA), and an extreme learning machine (ELM) are used to predict early CKD.…”