Predicting overall survivability in comorbidity of cancers: A data mining approach

The use of profit scoring as an alternative to credit scoring systems in peer-to-peer (P2P) lending

“…Explanation requires only cross validation whereas prediction requires intertemporal validation, which implies testing predictive results over time [35]. [36,37] and credit scoring [38,39]. CHAID is a recursive partitioning method that for regression-type problems relies on the F-square test to determine the best next split at each step.…”

A model to predict the survivability of cancer comorbidity through ensemble learning approach

“…Several prediction methods are examined by the means of train‐test split, and Gradient Boosting is selected as the predictor due to its superior performance. Given the outcome of the experiments run here, it is observed that in the proposed model, the accuracy, error, sensitivity, and specificity rates in predicting survival in cancer comorbidity is improved compared with the methods assessed by Zolbanin et al () in almost all cases, Table .…”

“…The sub‐categories: YR1973_2011.SEER9 YR1992_2011.SJ_LA_RG_AK YR2000_2011.CA_KY_LO_NJ_GA YR2005.LO_2ND_HALF The files, in ASCII file format: BREAST.TXT: Women breast cancer COLRECT.TXT: Colon and anus cancer DIGOTHR.TXT: Gastrointestinal cancer FEMGEN.TXT: Female genital cancer LYMYLEUK.TXT: Lymphoma of all sites and leukaemia MALEGEN.TXT: Men's genital system cancers RESPIR.TXT: Respiratory Cancer URINARY.TXT: Men urinary cancer OTHER.TXT: Other cases It should be noted that SEER data set applied in this study is that of 1973–2013, whereas in Zolbanin et al () the same data set applied is that of 1993‐2011. Regarding the difference in the dates of the two data sets, it is not possible to compare the results of this proposed method with Zolbanin et al in a direct manner.…”

“…Increased information about the cancer comorbidity enhances the power of the corresponding predictive models and assists the physicians to diagnose and treat the cancer. Proper identification, recording, and applying the patient comorbidity status can potentially reduce the costs of treatment and ultimately reduce the health‐related economic challenges (Zolbanin, Delen, & Zadeh, ).…”

“…By applying SEER cancer data, the researchers devised the two data sets applicable in predicting survival rates in comorbidity of breast and genital cancers in women and in specifying the survival rate of genital and urinary cancers in men. The results indicate that richer information about patient illnesses can improve the prediction power and assist physicians to better diagnosis and draw decisions (Zolbanin et al, ).…”