A New Modified Technique to Identify Outlier Values Using Fuzzy Clustering

Outliers within a dataset are data points that substantially differ from the rest of the data. These atypical data points can be attributed to a range of factors, such as errors in measurement, issues with data input, and natural variations in the data. Managing outliers is essential to ensure the integrity of statistical analyses and to avoid obtaining misleading results.These outliers can be observed at either the very high or very low points, possess the capacity to exert a notable effect on statistical measures such as the mean and variance. Many diagnostic techniques focus on affect centroids and distances between clusters to detect these abnormal points.In this particular study, fuzzy cluster techniques are primarily employed to identify outliers within the dataset. we proposed an alternative technique to detect outliers by using the concept and approach of standardization by using fuzzy cluster techniques in detection outliers in data set. The performance of the proposed method is compared with the others by using simulation.