Blockchain based Attack Detection on Machine Learning Algorithms for IoT\n based E-Health Applications

Abstract

The application of machine learning (ML) algorithms are massively scaling-up\ndue to rapid digitization and emergence of new tecnologies like Internet of\nThings (IoT). In today's digital era, we can find ML algorithms being applied\nin the areas of healthcare, IoT, engineering, finance and so on. However, all\nthese algorithms need to be trained in order to predict/solve a particular\nproblem. There is high possibility of tampering the training datasets and\nproduce biased results. Hence, in this article, we have proposed blockchain\nbased solution to secure the datasets generated from IoT devices for E-Health\napplications. The proposed blockchain based solution uses using private cloud\nto tackle the aforementioned issue. For evaluation, we have developed a system\nthat can be used by dataset owners to secure their data.\n