By Jing Gao
Outlier (or anomaly) detection is a really large box which has been studied within the context of a giant variety of study components like facts, facts mining, sensor networks, environmental technological know-how, dispensed platforms, spatio-temporal mining, and so forth. preliminary study in outlier detection considering time series-based outliers (in statistics). because then, outlier detection has been studied on a wide number of facts forms together with high-dimensional facts, doubtful info, flow info, community info, time sequence information, spatial info, and spatio-temporal information. whereas there were many tutorials and surveys for basic outlier detection, we specialise in outlier detection for temporal info during this booklet. quite a few purposes generate temporal datasets. for instance, in our lifestyle, several types of documents like credits, group of workers, monetary, judicial, clinical, etc., are all temporal. This stresses the necessity for an equipped and precise examine of outliers with appreciate to such temporal facts. some time past decade, there was loads of learn on numerous varieties of temporal facts together with consecutive info snapshots, sequence of knowledge snapshots and knowledge streams. along with the preliminary paintings on time sequence, researchers have interested by wealthy types of info together with a number of facts streams, spatio-temporal info, community info, neighborhood distribution info, etc.
Compared to common outlier detection, thoughts for temporal outlier detection are very diverse. during this booklet, we are going to current an geared up photograph of either contemporary and previous study in temporal outlier detection. we begin with the fundamentals after which ramp up the reader to the most principles in cutting-edge outlier detection suggestions. We inspire the significance of temporal outlier detection and short the demanding situations past traditional outlier detection. Then, we record down a taxonomy of proposed concepts for temporal outlier detection. Such thoughts extensively contain statistical options (like AR versions, Markov types, histograms, neural networks), distance- and density-based techniques, grouping-based methods (clustering, group detection), network-based techniques, and spatio-temporal outlier detection methods. We summarize by means of proposing a large choice of functions the place temporal outlier detection concepts were utilized to find fascinating outliers.
Table of Contents: Preface / Acknowledgments / determine credit / creation and demanding situations / Outlier Detection for Time sequence and information Sequences / Outlier Detection for info Streams / Outlier Detection for dispensed facts Streams / Outlier Detection for Spatio-Temporal information / Outlier Detection for Temporal community information / functions of Outlier Detection for Temporal information / Conclusions and study instructions / Bibliography / Authors' Biographies
Read or Download Outlier Detection for Temporal Data (Synthesis Lectures on Data Mining and Knowledge Discovery) PDF
Best Data Mining books
Freemium Economics provides a pragmatic, instructive method of effectively enforcing the freemium version into your software program items through development analytics into product layout from the earliest levels of improvement. Your freemium product generates huge volumes of information, yet utilizing that info to maximise conversion, develop retention, and carry profit could be hard if you happen to do not absolutely comprehend the impression that small adjustments could have on profit.
Placed Predictive Analytics into motion research the fundamentals of Predictive research and information Mining via a simple to appreciate conceptual framework and instantly perform the suggestions discovered utilizing the open resource RapidMiner software. even if you're fresh to information Mining or engaged on your 10th undertaking, this e-book will allow you to research info, discover hidden styles and relationships to assist vital judgements and predictions.
Information warehousing is among the most well-liked company subject matters, and there’s extra to knowing facts warehousing applied sciences than you could imagine. discover the fundamentals of knowledge warehousing and the way it enables facts mining and company intelligence with information Warehousing For Dummies, second version. facts is maybe your company’s most vital asset, so your info warehouse should still serve your wishes.
Info Mining in Finance offers a entire evaluation of significant algorithmic techniques to predictive info mining, together with statistical, neural networks, ruled-based, decision-tree, and fuzzy-logic tools, after which examines the suitability of those ways to monetary facts mining. The ebook focuses particularly on relational info mining (RDM), that is a studying procedure in a position to study extra expressive ideas than different symbolic techniques.
Extra info for Outlier Detection for Temporal Data (Synthesis Lectures on Data Mining and Knowledge Discovery)