Review of Immunotherapy Classification: Application Domains, Datasets, Algorithms and Software Tools from Machine Learning Perspective

Abstract

Immunotherapy treatments can be essential sometimes and a waste of valuable resources in other cases, depending on the diagnosis results. Therefore, researchers in immunotherapy need to be updated on the current status of research by exploring: application domains e.g warts, datasets e.g immunotherapy, classifiers or algorithms e.g kNN, software tools e.g. python, and publications. The objective was to study the immunotherapy related published literature, from a supervised machine learning perspective. In addition, to reproduce research papers implementations of Random Forest and kNN among other algorithms. To find gaps and challenges both in publications and practical work, which may be the basis for further research. Immunotherapy, diabetes, cryotherapy, exasens data and one unbalanced dataset are explored. Random Forest performed better. The results are compared with published literature. To address the found gaps in further research: novel experiments, unbalanced studies, focus on effectiveness and a new algorithm or classifier are suggested.

Keywords

• immunotherapy
• cryotherapy
• machine learning
• classification
• random forest
• warts