The process of mining comprises of supervised learning and unsupervised learning. It includes various approaches out of which data classification is one of the beneficial and constructive methods. This paper explores the effective functioning of the whole process. There are several cases in classification where the important data is missed during the process. It can hence be concluded that the process of mining is greatly affected by the absence of such kind of data. The process of extraction of unknown and new data from the huge dataset that has been left during the classification process is known as novelty detection. It aims to cover the most commonly used ideas in modeling the novelty data, to classify them as well as to provide useful methods which need to be used in order to identify novelty data for temporal data time series. Novelty detection generally focuses on the identification of shapes or patterns, space, density, distance and learning models. This paper mainly aims to identify the important data which is typically missed during the process of training. This research explains the overall concept to detect the novelty data that is used in the taxonomy of level set methods.
A conceptual paper on novelty detection for temporal data using level set methods
2017-04-01
309436 byte
Conference paper
Electronic Resource
English
Temporal video segmentation by event detection: A novelty detection approach
British Library Online Contents | 2014
|AIAA | 2016
|Structural fault detection using a novelty measure (sv960747)
Online Contents | 1997
|