New hybrid ensemble method for anomaly detection in data science

Anomaly detection is a significant research area in data science. used to find unusual points or uncommon events streams. It gaining popularity not only the business world but also different of other fields, such as cyber security, fraud for financial systems, and healthcare. Detecting anomalies could be useful new knowledge data. This study aims build an effective model protect from these anomalies. We propose hyper ensemble machine learning method that combines predictions two methodologies outcomes isolation forest-k-means random forest using voting majority. Several available datasets, including KDD Cup-99, Credit Card, Wisconsin Prognosis Breast Cancer (WPBC), Forest Cover, Pima, were evaluate proposed method. The experimental results exhibit our gives highest realization terms receiver operating characteristic performance, accuracy, precision, recall. Our approach more efficient detecting than approaches. accuracy rate achieved 99.9%, compared without method, which achieves 97%.