Paper Title
Outlier Detection Using Machine Learning

Abstract
Outlier detection is essentially the process of finding the data points that differ significantly from the rest of the data. The sources of outliers in a data sample may be varied, and the causes of such outliers range from fraud, data collection errors, and inherent data variability. Outlier detection is the prime task in any machine learning project, be it fraud detection, medical diagnosis, or network security. Research done till date brings to us various techniques to detect these outliers, with their respective positives and negatives. In our review paper, we humbly attempt to provide an overview of this field's research, from the earliest stages to the latest developments. We attempt to discuss various categories of machine learning methods and provide reasoning for their usability and scope of improvement. Towards the conclusion, a hybrid method has been proposed to bypass the drawbacks of most of the methods. We also discuss the further scope of outlier detection in machine learning where deep learning and semi-supervised methods revolutionize the subject. This review paper will employ valuable insights from the literature to provide a comprehensive overview of the state-of-the-art in outlier detection using machine learning. Keywords - Outlier, Clustering, Ensemble Learning, Local Outlier Factor, Deep Outlier Detection, Hyper-parameters