Perception-based approach to time series data mining

Batyrshin, Ildar; Sheremetov, Leonid

doi:10.1016/j.asoc.2007.02.020

Cited by 44 publications

(19 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Related to the present research, Batyrshin and Sheremetov (2008) mention that extracting meaningful data from databases can be useful for management and decision purposes. The development of text mining techniques to extract information from a time series database can contribute on this case.…”

Section: Text Miningmentioning

confidence: 99%

Aircraft interior failure pattern recognition utilizing text mining and neural networks

et al. 2011

View full text Add to dashboard Cite

Being more competitive is routine in the aeronautical sector. Airline competitiveness is affected by such factors as time, price, reliability, availability, safety, technology, quality, and information management. To remain competitive, airlines must promptly identify and correct failures found in their fleet. This study aims at reducing the time spent on identifying and correcting such failures logged. Utilizing Text Mining techniques during the pre-processing phase, our study processes an extensive database of events from commercial regional jets. The result is a unique list of keywords that describes each reported failure. Later, an Artificial Neural Network (ANN) identifies and classifies failure patterns, yielding a respective disposition for 742 J Intell Inf Syst (2012) 38:741-766 a given failure pattern. Approximately five years of historical data was used to build and validate the present model. Results obtained were promising.

show abstract

Section: Text Miningmentioning

confidence: 99%

Aircraft interior failure pattern recognition utilizing text mining and neural networks

et al. 2011

View full text Add to dashboard Cite

show abstract

“…14 Batyrshin and a The query language employed is not relevant for the purpose of our work. coworker 15,16 have also work with time series data mining and representation of quantitative dependencies.…”

Section: Introductionmentioning

confidence: 99%

Linguistic query answering on data cubes with time dimension

Castillo-Ortega

Marín

Sánchez

2011

Int. J. Intell. Syst.

View full text Add to dashboard Cite

In this paper, we propose a methodology for providing linguistic answers to queries involving the comparison of time series obtained from data cubes with time dimension. Time series related to events which are interesting for the user are obtained by querying data cubes using OnLine Analytical Processing (OLAP) operations on the time dimension. The comparison of these query results can be summarized so that an appropriate short linguistic description of the series is provided to the user. Our approach is based on linguistically quantified statements and pointwise definitions of the degree and sign of local change. Our linguistic summaries are well suited to be included in an interface layer of a data warehouse system, improving the quality of humanmachine interaction and the understandability of the results. C 2011 Wiley Periodicals, Inc.

show abstract

“…This technique can be extended to general time series data if nominal TS features are defined as events. The concept of an episode emphasises together with the importance of combinations of various events, that this conforms to the manner in which humans often perceive temporal events (Batyrshin & Sheremetov 2008). Despite this extensive research on episode mining considers the discovery of frequent serial episodes as the fundamental problem (Mannila et al 1997, Laxman et al 2007, as these frequent episodes are believed to have a higher importance than infrequent episodes.…”

Section: Figure 2: a Diagram Of The Blast Furnace In The Iron-making mentioning

confidence: 99%

An Event Group Based Classification Framework for Multi-variate Sequential Data

Sun

Stirling

2017

AJIS

View full text Add to dashboard Cite

Decision tree algorithms were not traditionally considered for sequential data classification, mostly because feature generation needs to be integrated with the modelling procedure in order to avoid a localisation problem. This paper presents an Event Group Based Classification (EGBC) framework that utilises an X-of-N (XoN) decision tree algorithm to avoid the feature generation issue during the classification on sequential data. In this method, features are generated independently based on the characteristics of the sequential data. Subsequently an XoN decision tree is utilised to select and aggregate useful features from various temporal and other dimensions (as event groups) for optimised classification. This leads the EGBC framework to be adaptive to sequential data of differing dimensions, robust to missing data and accommodating to either numeric or nominal data types. The comparatively improved outcomes from applying this method are demonstrated on two distinct areas -a text based language identification task, as well as a honeybee dance behaviour classification problem. A further motivating industrial problem -hot metal temperature prediction, is further considered with the EGBC framework in order to address significant real-world demands.

show abstract

Perception-based approach to time series data mining

Cited by 44 publications

References 16 publications

Aircraft interior failure pattern recognition utilizing text mining and neural networks

Aircraft interior failure pattern recognition utilizing text mining and neural networks

Linguistic query answering on data cubes with time dimension

An Event Group Based Classification Framework for Multi-variate Sequential Data

Contact Info

Product

Resources

About