A blockchain network's economics and user confidence can be seriously harmed by fraud. Consensus algorithms like proof of work and proof of stake can verify the legitimacy of a transaction but not the identity of the people who are conducting or verifying it. On a blockchain network, fraud can still occur, as a result of this. One approach to fighting fraud is to make use of machine learning techniques. There are two types of machine learning: supervised and unsupervised. We use a variety of supervised machine learning techniques in this study to distinguish between legitimate and fraudulent purchases. We also compare decision trees, Naive Bayes, logistic regression, multilayer perceptron, and other supervised machine learning techniques in detail for this challenge.

Incursion finding is a process of identifying and responding malicious activity. Wireless sensor webs consisting bulk of ‘sensors’ are useful to integrate data in variety of environment. The basic sensors are simple and have limited power supplies. Heterogeneous wireless sensor webs are better scalable and lower overall cost than homogeneous sensor webs. In this paper, we are improving the lifetime of wireless web and we present a survey of various energy efficient techniques in a various wireless sensor web. It is important to improving wireless web because sensor nodes in wireless webs are constrained by limited energy.