ADAM Optimizer and CATEGORICAL CROSSENTROPY Loss Function-Based CNN Method for Diagnosing Colorectal Cancer
Joyita Ghosh, Subir Gupta
Abstract
Over the previous four decades, significant advancements have been made in the field of medical science. During this time in history, not only were new ways to diagnose diseases and make medicines, but the root causes of a number of illnesses were also found. There have been many breakthroughs, but as long as humans have a biological vulnerability to diseases like cancer, this will remain a challenging challenge to overcome. Cancer is the second leading cause of death worldwide, killing one out of every six people who are diagnosed with the disease. Even though there are many different types of colorectal cancer, the ones that affect the colon are by far the most prevalent and deadly. However, if the disease is diagnosed at a relatively early stage, there is a substantial increase in the likelihood that the patient will live. Machine learning makes it possible to evaluate more circumstances in less time and for less money, paving the way for the creation of an automatic technique for the diagnosis of cancer. This research, which CNN paid to promote, presents a new paradigm for identifying colorectal cancer at a more advanced stage. When it comes to recognizing malignant illnesses, the proposed framework offers a maximum accuracy of AAA%, according to the findings. Throughout our investigation, the study used the ADAM optimizer and the categorical cross-entropy loss function. Then, the study compared our findings to those obtained using alternative optimizers and loss functions.