<p>Cancer is a diverse disease brought on by the aberrant development of a cell. Cancer prognosis has become a priority in cancer research as it helps medical experts to plan treatment strategies. As a result, Artificial Intelligence-based learning approaches have been used to model the development and treatment of cancer. In cancer research, several such techniques are being commonly used to build prediction models, leading to an efficient and reliable verdict. We examined the most recent models to highlight the role of predictive models in cancer research. Using a combination of keywords like machine learning, deep learning, cancer prognosis, cancer prediction, and cancer survivability, we found articles published in scientific databases between 2010 and 2021. Our study carries multiples investigations and presents a summary of automated learning methods used in cancer predictive modeling in this article. Although it is evident that machine/deep learning-based models enhance our ability in the early prediction of cancer diagnosis, an adequate amount of affirmation is needed to make sure that such predictive models can be introduced in routine medical practices. The cancer research field is of high impact, and investigations on cancer prediction models should be encompassed for medical decision support of both practitioners and patients.</p>