A Model to Predict Survival among Pancreatic Cancer Patients using Supervised Machine Learning
Abstract: Cancer is a diverse illness with uncontrolled cell division and the propensity to spread. It has continuous cell division, no programmed cell death, tissue invasion, and blood vessel formation at its location. Oncologists can use cancer survival rates to determine a patient's prognosis and treatment approach. Age, race, cancer site, stage, grade, number of tumors, kind of cancer; and other factors affect patient survival. This study focused on Pancreatic Cancer patient survival prediction using CNN on a dataset that with 7558 labelled images.The model predicted cancer survival with a maximum score of 1 which is the highest possible score on all the metrics used. However, the confusion matrix of the model indicates that the model made 10 false positives and 9 false negatives making a total of 19 false predictions.The model's training and validation loss were 0.02 and 0.03. These loss values are low and converge to zero, indicating effective model prediction and dependability.
Keywords: Pancreatic Cancer, Supervised Machine Learning
How to Cite?: Simbarashe N. Manyetu, Monika Gondo, "A Model to Predict Survival among Pancreatic Cancer Patients using Supervised Machine Learning", Volume 11 Issue 12, December 2022, International Journal of Science and Research (IJSR), Pages: 283-286, https://www.ijsr.net/getabstract.php?paperid=SR221204211131, DOI: https://dx.doi.org/10.21275/SR221204211131