Effectiveness and survival status of patients through robotic surgery in healthcare sector using machine learning algorithm
Authors
Thakur, Laxman Narayan
Issue Date
2020
Degree
MSc in Data Analytics
Publisher
Dublin Business School
Rights holder
Rights
Items in eSource are protected by copyright. Previously published items are made available in accordance with the copyright policy of the publisher/copyright holder.
Abstract
Medicine has experienced greater scientific and technological advances in last few years than in the rest of human history. Ever since computer technology entered the operating room, surgery has gone through one of the greatest changes throughout medical life. The major potential of robotic surgery are precision and miniaturization. The exposure of Robotic surgery using machine learning techniques is used to predict survival status of patients with regards to classification of multiple feedback from people and effectiveness to people in surgeries can help to improve betterment of life in healthcare sector. This research includes data in the year of 2015 at the California Research development and health care service which has information about patients in terms of Robotic surgery. We used almost five machine algorithms mainly Random Forrest, MLP neural network, Naïve Bayes, KNN, SVM. Random Forrest and Naïve Bayes are best fit and ability to learn from data so accurately that gives better insights and results in terms of Usefulness of Robotic surgery in certain low dimension space. MLP neural network gives the better prediction and performance when survival status of patients is concerned. The overall execution of these 3 models are measured by confusion matrix and accuracy of each model. The accuracy given by these models conclude that MLP neural network (97%) , Naïve Bayes(95%) has best figures followed by Random Forrest(92%). Therefore, MLP Neural network with Random Forrest and Naïve Bayes are most suitable and optimised models for this specific study of prediction of Robotic surgery in respective cases.