OUTLIER DETECTION IN CATEGORICAL DATA

Bücher

Angebote / Angebote:

Datasets are characterized by the properties of the majority of the data objects in it. There are a few data objects whose characteristics are not similar to the mainstream characteristics of the data objects in a dataset. These data objects may contain valuable information and are called outliers. Outlier detection is an important concept in data mining due to its application in a wide range of ¿elds. Outlier detection refers to the problem of ¿nding hidden observations with vital information whose properties are not similar to the properties of the mainstream observations in the dataset. Outlier detection was not an interesting research area till the last decade. In recent years, outlier detection has been investigated by a number of researchers because of its importance in a wide range of application areas and di¿erent techniques have been developed for ¿nding outliers in various domains. Outliers are also called anomalies in the literature. Depending on the application domains and context, they are also referred to as exceptions, errors, discordant observations, noises, faults, defects, aberrations, novelties, peculiarities or contaminants. Earlier, outlier detection was a research topic in Statistics. Nowadays, it is a research area in many branches of science including Data Mining and Machine learning.