Deep Learning and Multiclass Machine Learning Classifier Approach for Predicting Primary Tumors

Abstract

Deep Learning (DL) and Machine Learning (ML) have the great prospect to play a significant role in the medical field in disease prediction. The tumor or cancer is one of the major health issues that each nation is currently dealing with, and it is the topic of this essay. The prediction of unidentified primary tumors in the dataset is delineated in this paper. Given that it provides significantly higher accuracy than binary classifiers, different multiclass classifier such as K-Nearest Neighbor (KNN), CatBoost Classifier, Random Forest Classifier, Gradient Boosting Classifier, Light Gradient Boosting Machine, Ada Boost Classifier, Decision Tree Classifier, SVM - Linear Kernel, Naive Bayes and Deep neural networks (DNN1, DNN2, and DNN3) are used to categorize multiclass datasets available in the UCI machine learning repository. Among the stated machine learning classifiers, the k-Nearest Neighbor (KNN) had the highest classification accuracy of 92.92%. The three layer deep neural network (DNN2), among deep learning techniques, had produced the best accuracy of 97.66% using the chosen features as input. The gathered results from this work showed that deep neural networks outperformed machine learning techniques.