Towards a Framework for Change Detection in Data Sets

Böttcher, Mirko; Nauck, Detlef; Ruta, Dymitr; Spott, Martin

doi:10.1007/978-1-84628-663-6_9

Cited by 10 publications

(12 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Moreover, such a temporal moving window approach performs similar to age dependent weighting approaches -the alternative method to learn decision trees in the presence of concept drift -in case of smooth, non-abrupt concept drift [11,10]. Such a type of concept drift is present in our data as we know from previous studies on change mining carried out on the same data set [2]. 4 Classification accuracy for two different attribute evaluation measures in several consecutive time periods T .…”

Section: Experimental Evaluationmentioning

confidence: 86%

“…The information gain history of the attribute A (1) is apart from noise stable whereas the information gain history of A (2) shows an upward trend. Furthermore, it can be seen that for the vast majority of time periods T = 1, .…”

Section: Basic Ideamentioning

confidence: 92%

“…,15 attribute A (1) has more predictive power and would therefore been chosen as the split attribute. However, due to the observed upward trend in the information gain of A (2) both histories will intersect and A (2) will become the split attribute in the near future. …”

Section: Basic Ideamentioning

confidence: 93%

See 2 more Smart Citations

An Algorithm for Anticipating Future Decision Trees from Concept-Drifting Data

Böttcher

Spott

Kruse

2009

Research and Development in Intelligent Systems XXV

Self Cite

View full text Add to dashboard Cite

Concept-Drift is an important topic in practical data mining, since it is reality in most business applications. Whenever a mining model is used in an application it is already outdated since the world has changed since the model induction.The solution is to predict the drift of a model and derive a future model based on such a prediction. One way would be to simulate future data and derive a model from it, but this is typically not feasible. Instead we suggest to predict the values of the measures that drive model induction. In particular, we propose to predict the future values of attribute selection measures and class label distribution for the induction of decision trees. We give an example of how concept drift is reflected in the trend of these measures and that the resulting decision trees perform considerably better than the ones produced by existing approaches.

show abstract

Section: Experimental Evaluationmentioning

confidence: 86%

Section: Basic Ideamentioning

confidence: 92%

See 1 more Smart Citation

An Algorithm for Anticipating Future Decision Trees from Concept-Drifting Data

Böttcher

Spott

Kruse

2009

Research and Development in Intelligent Systems XXV

Self Cite

View full text Add to dashboard Cite

show abstract

“…we introduced a new mining problem of partial periodic pattern with random replacement. [9] 4. Existing Chatterbot system:…”

Section: Info Miner Algorithmmentioning

confidence: 99%

“…In the case of frequent itemsets by exploiting generalized itemsets to represent patterns that become unusual with respect to the support threshold, and therefore they are no longer extracted, at a specific point. [3], [8], [9].This type of algorithm is used to discovering basic changes [6], [13], [14]). Consider the itemsets {Paint, Bihar} and {Jacket, Paris}.…”

Section: Introductionmentioning

confidence: 99%

Survey on Discovering rapid Changing Patterns In the presence of Hierarchy

Priya¹,

Nawathe²

2014

IOSRJCE

View full text Add to dashboard Cite

Frequent item set mining is used that focuses on find out recurrent correlations in the data. Change mining, it focuses on frequent itemsets, focuses on important changes in the set of mined itemsets from one point in time period to another. The finding of frequent generalized itemsets, One dynamic pattern, the history generalized pattern ,that represents the development of an itemset in successive time periods, by accounting the information about its recurrent generalizations characterized by minimal redundancy some time it becomes infrequent. Higen mining, The higen miner, that focuses on avoiding itemset mining followed by postprocessing by developing a support-driven itemset generalization .To focus the attention on the minimally redundant recurrent generalizations and reduce the amount of the generated patterns, the finding a subset of higens, namely the nonredundant higens,. Tests do on both real and synthetic datasets show the competence and the effectiveness [1] .

show abstract

Contrast and change mining

Boettcher

2011

WIREs Data Min & Knowl

View full text Add to dashboard Cite

Because the world with its markets, innovations, and customers is changing faster than ever before, the key to survival for businesses is the ability to detect, assess, and respond to changing conditions timely and intelligently. Understanding changes and reacting to or acting upon them therefore become a strategic issue not only for companies but also in many other domains. The corresponding need for knowledge has been answered by data mining research by proposing a multitude of methods for analyzing different aspects of change. This article provides an overview of recent works on methods for change analysis, thereby focusing on contrast mining and change mining, the two emerging subfields of contemporary data mining research.

show abstract

Towards a Framework for Change Detection in Data Sets

Cited by 10 publications

References 12 publications

An Algorithm for Anticipating Future Decision Trees from Concept-Drifting Data

An Algorithm for Anticipating Future Decision Trees from Concept-Drifting Data

Survey on Discovering rapid Changing Patterns In the presence of Hierarchy

Contrast and change mining

Contact Info

Product

Resources

About