Leveraging Blockchain Driven Stacked Model For Ransomware Detection In Bitcoin Transactions

Abstract

To improve interoperability and privacy among system users, smart grids must allow for the sharing of data and information. Traditional cloud-based data interchange techniques, on the other hand, have been centralised on a single platform run by a dependable third party, which has resulted in single points of failure, inadequate data protection, and unrestricted access. Blockchain technology has been suggested as a decentralised and secure platform for data exchange within smart grids to overcome these problems. This innovative platform offers solutions to important issues like privacy, scalability, and user ownership and enables safe data trading between users without ownership loss. Participants can access data programmatically with the help of blockchain-based smart contracts while guaranteeing that every interaction is verified and documented by other users of the tamper-resistant blockchain network. In this regard, a new model for forecasting S&P500 volatility has been put out employing a variety of machine learning methods, including Gradient Descent Boosting, Random Forest, Support Vector Machine, and Artificial Neural Network. In order to increase the predictability of the predictions, these algorithms have been stacked. Resilient K-NN and FLR have been combined in the suggested stacking model for prediction, and it has demonstrated classification accuracy of 97.41% and 88.27%, respectively. These experimental findings suggest that it is advantageous to look into the use of PART for tasks involving predictive modelling in Ransomware investigations.