http://localhost:8080/xmlui/handle/123456789/999 2017-18 : Research Article/Paper 2026-06-23T06:31:47Z http://localhost:8080/xmlui/handle/123456789/1004 Title: A Review on Outlier Detection Techniques Authors: Gupta, D. R.; Kamlapurkar, S. M. Abstract: The outlier is unexpected behavior of data. Outlier detection is important in various domains like fraud detection, intrusion detection, activity monitoring, etc. Data is generated continuously on large scale in many applications. There is need to detect outlier from static data as well as streaming data. There are basic two types of outlier: global outlier and local outlier. This work aims to study various local and global outlier detection techniques for static and streaming data. The works also focuses on various local and global outlier detection techniques which are efficient in terms of time and memory. 2017-12-01T00:00:00Z http://localhost:8080/xmlui/handle/123456789/1003 Title: A Review on Image Annotation Generation Authors: Aher, A. S.; Kamlapurkar, S. M. Abstract: Image annotation, tagging, semantic descriptors are used to search an image from large image dataset. To search similar image, low level feature matching is used. In text based image search process image tags, annotations or image semantics are used. Annotation can be generated using manual, semi-automatic or automatic process. The manual and semiautomatic annotations are time consuming. In automatic image annotation, tag is automatically assigned to an image. There are various techniques for annotation generation. This work aims to study various image annotations generation, image matching and image searching techniques based on execution strategies and efficiency evaluation. 2017-12-01T00:00:00Z http://localhost:8080/xmlui/handle/123456789/1002 Title: Analysis of Graph Clustering Method Authors: Boraste, P. S.; Kanmlapurkar, S. M. Abstract: Network data clustering has vital importance in various domains such as social network analysis, epidemiology, World Wide Web analysis, etc. The clustering technique derives underlying structures present in the graph. Along with the cluster creation, vertices classification is also an important task. To detect hubs and outlier is very important task in graph mining. There are various graph clustering algorithm such as graph partitioning, density based, modularity based, etc. This work aims to study various techniques of graph data clustering 2017-12-01T00:00:00Z http://localhost:8080/xmlui/handle/123456789/1001 Title: A Review on Feature Subset Creation Strategies Authors: Patil, P. P.; Banait, S. S. Abstract: To reduce computational overhead in processing high dimensional dataset, dimensionality reduction is important mechanism to remove redundant and unused attributes from dataset in data analysis phase. Feature selection and feature extraction are two techniques in dimensionality reductions. This work aims to study various techniques involved in feature subset generation and reduction of data set size, its efficiency in terms of execution time and quality analysis parameters. 2017-12-01T00:00:00Z