MOCA-I: Discovering Rules and Guiding Decision Maker in the Context of Partial Classification in Large and Imbalanced Datasets

Jacques, Julien; Taillard, Julien; Delerue, David; Jourdan, Laëtitia; Dhaenens, Clarisse

doi:10.1007/978-3-642-44973-4_5

Cited by 4 publications

(3 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The total number of RVQA data confidentiality violation detections was evaluated using confidence and sensitivity metrics [54][55][56]. Confidence (precision) is the percentage of the detected violations from the overall detection result that are correct (when compared to the known data confidentiality violations).…”

Section: Results Validation Using Evaluation Metricsmentioning

confidence: 99%

Improving Agent Quality in Dynamic Smart Cities by Implementing an Agent Quality Management Framework

2019

View full text Add to dashboard Cite

It is critical for quality requirements, such as trust, privacy, and confidentiality, to be fulfilled during the execution of smart city applications. In this study, smart city applications were modeled as agent systems composed of many agents, each with its own privacy and confidentiality properties. Violations of those properties may occur during execution due to the dynamic of agent behavior, decision-making capabilities, and social activities. In this research, a framework called Agent Quality Management was proposed and implemented to manage agent quality in agent systems. This paper demonstrates the effectiveness of the approach by applying it toward a smart city application called a crowdsourced navigation system to verify and assess agent data confidentiality. The AnyLogic Agent-Based Modeling tool was used to model and conduct the experiments. The experiments showed that the framework helped to improve the detection of agent quality violations in a dynamic smart city application. The results can be further analyzed using advanced data analytic approach to reduce future violations and improve data confidentiality in a smart city environment.

show abstract

Section: Results Validation Using Evaluation Metricsmentioning

confidence: 99%

Improving Agent Quality in Dynamic Smart Cities by Implementing an Agent Quality Management Framework

2019

View full text Add to dashboard Cite

show abstract

“…Each study is realized with the following default parameters, which are the default parameters used in a previous version of MOCA-I [43]:…”

Section: Resultsmentioning

confidence: 99%

Conception of a dominance-based multi-objective local search in the context of classification rule mining in large and imbalanced data sets

Jacques

Taillard²,

Delerue³

et al. 2015

Applied Soft Computing

Self Cite

View full text Add to dashboard Cite

“…In this section we will see which encoding and neighborhood we use in the partial classification rule mining problem, and which algorithms are adapted to deal with this problem in a single-objective way and then in a multi-objective way. We use the encoding and the neighborhood used in MOCA-I algorithm detailed in [16]. Each solution is represented using the Pittsburgh encoding, where each solution is a variablelength set of rules.…”

Section: Encoding and Neighborhoodmentioning

confidence: 99%

The benefits of using multi-objectivization for mining pittsburgh partial classification rules in imbalanced and discrete data

Jacques¹,

Taillard²,

Delerue³

et al. 2013

Proceedings of the 15th Annual Conference on Genetic and Evolutionary Computation

Self Cite

View full text Add to dashboard Cite

A large number of rule interestingness measures have been used as objectives in multi-objective classification rule mining algorithms. Aggregation or Pareto dominance are commonly used to deal with these multiple objectives. This paper compares these approaches on a partial classification problem over discrete and imbalanced data. After performing a Principal Component Analysis (PCA) to select candidate objectives and find conflictive ones, the two approaches are evaluated. The Pareto dominance-based approach is implemented as a dominance-based local search (DMLS) algorithm using confidence and sensitivity as objectives, while the other is implemented as a single-objective hill climbing using F-Measure as an objective, which combines confidence and sensitivity. Results shows that the dominance-based approach obtains statistically better results than the singleobjective approach.

show abstract

MOCA-I: Discovering Rules and Guiding Decision Maker in the Context of Partial Classification in Large and Imbalanced Datasets

Cited by 4 publications

References 19 publications

Improving Agent Quality in Dynamic Smart Cities by Implementing an Agent Quality Management Framework

Improving Agent Quality in Dynamic Smart Cities by Implementing an Agent Quality Management Framework

Conception of a dominance-based multi-objective local search in the context of classification rule mining in large and imbalanced data sets

The benefits of using multi-objectivization for mining pittsburgh partial classification rules in imbalanced and discrete data

Contact Info

Product

Resources

About