A Survey on Machine Learning Techniques

Abstract

Machine learning is one of the most popular field in the computer science having different types of techniques such as supervised learning, unsupervised learning, reinforcement learning and the various techniques which are lying under them, so in order to understand these different machine learning techniques a survey on these machine learning techniques has been done and tried to explain these few techniques. The machine learning techniques try to understand the different data sets which are given to the machine. The data which comes inside can be divided into two types i.e. labelled data and the unlabeled data. These have to tackle both of the data. Those techniques have been looked upon as well. Then the concept of outlier comes into picture. Outlier Detection is one of the major issues in Data Mining; to find an outlier from a group of patterns is a famous problem in data mining. A pattern which is dissimilar from all the remaining patterns is an outlier in the dataset. Earlier outliers were known as noisy data, now it has become very difficult in different areas of research. Finding an outlier is useful in detecting the data which can’t be predicted and that which can’t be identified. A number of surveys, research and review articles cover outlier detection techniques in great details. The paper discusses and it tries to explain some of the techniques which can help us in identifying or detecting the observation which show such kind of abnormal behavior, and in technical terms called as outlier detection techniques.