AI study on Tumor Identification

Abstract

Artificial Intelligence (AI), specifically machine learning and deep learning, have become a hot topic in recent years. Everyone is excited and curious about how AI can shape our society and change our future. Indeed, AI is already emerged in our normal life and will be applied in various industries gradually. According to a report by Research and Markets, the global Artificial Intelligence market size is expected to reach almost $170,000 million in 2025, from $4,000 million in 2016. The demand for AI solution of market will keep increasing in terms of revenue brings to the company. AI is a powerful tool which can act like a human brain. It can do the "thinking" process such as predicting, analysing, planning, making decision, etc. Although the development of AI industry is still progressing and not completely mature, we can take it as an assistance to enhance efficiency and human decision making. The one of promising and potential usages is in healthcare. Google has already launched an AI which can predict whether you are at risk of heart disease by looking at people's eyes while Stanford University designed an AI algorithm to detect skin cancers which performs same level as dermatologists. It is amazing that AI can support human to provide efficient and qualified service, by diagnosing and predicting disease of patients as soon as possible. In this project, we will investigate the implementation of the cancer prediction, starting from the data analysis to models comparison. Look into the black box of the models and have better understanding of their decision. After all, we will get an optimal solution which give out the best prediction score base on different criteria.