Volume 7 Issue 3 - September 2016

  • 1. An experimental analysis of outliers detection on static exaustive datasets.

    Authors : Raghav Purankar, Pragati Patil

    Pages : 319-325

    DOI : http://dx.doi.org/10.21172/1.73.544

    Keywords : Cluster based Approach, Data Mining, Distance based approach, Outlier Detection, UCI Repository.

    Abstract :

    Detecting Outlier and clustering methodologies are an important branch of data mining, by combining the two technologies can improve the data mining significance. Exhaustive data sets by experimental results ensures that the algorithm will improve the efficiency of outlier detection. The outliers may be instances of error or indicate different behavior. The task of outlier detection primarily focused on identifying such outliers so as to improve the data analysis and then find out only interesting and useful information about unusual events within number of application domains. Finding outliers in a group of patterns is a very well-known issue in the data mining. The principle of outlier detection depend on the threshold value. Threshold is generally provided by user. In proposed approach, two methods cluster based approach and distance based approach are applied individually over static data sets to efficiently find the outlier from the data set. The exhaustive datasets can be downloaded from the UCI machine learning repositories. The experimental results of real exhaustive dataset demonstrate that proposed method takes lesser cost in computation and gives better performance than the traditional methods.

    Citing this Journal Article :

    Raghav Purankar, Pragati Patil, "An experimental analysis of outliers detection on static exaustive datasets.", https://www.ijltet.org/journal_details.php?id=907&j_id=3356, Volume 7 Issue 3 - September 2016, 319-325, #ijltetorg