A Model to Predict Cancer Comorbid Patient’s Survivability Based on Improved Chi2 Model
作者单位：Key Laboratory of Measurement and Control of CSE Ministry of Education School of Automation Southeast University
会议名称：《第三十九届中国控制会议》
会议日期：2020年
学科分类：12[管理学] 1201[管理学-管理科学与工程(可授管理学、工学学位)] 1002[医学-临床医学] 081104[工学-模式识别与智能系统] 08[工学] 0835[工学-软件工程] 100214[医学-肿瘤学] 0811[工学-控制科学与工程] 0812[工学-计算机科学与技术（可授工学、理学学位）] 10[医学]
关 键 词：Cancer Comorbidity Survival Prediction LightGBM Chi2 model Data Balancing
摘 要：Modeling of Cancer comorbid patients’ survival probability has theoretical significance and practical needs. Cancer survivability prediction may provide advices for clinical decisions and personalized medicine. Surveillance, Epidemiology, and End Results Program（SEER） provides large data sets to be analyzed with machine learning methods. In this study, male and female cancer comorbid cases（male-genital and urinal for men and breast and female-genital for women） are identified and labeled from the SEER database;the data set is then processed with improved Chi2 test based feature selection and random undersampling based data-balancing;lightGBM is adopted as the classifier. The results indicate that the proposed method is effective and generalizable.