CASEP2: Hybrid Case-Based Reasoning System for Sequence Processing

Zehraoui, Farida; Kanawati, Rushed; Salotti, Sylvie

doi:10.1007/978-3-540-28631-8_33

Cited by 7 publications

(6 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Next, we propose a hybrid neuro-CBR system CASEP2 [73,74] where temporal aspect of data is taken into account by using a new case modeling as well as an adequate neural network. A maintenance strategy is also proposed in order to process the large amount of processed raw noisy data (i.e.…”

Section: Applying Neural Network To User Behavior Modelingmentioning

confidence: 99%

“…This is clearly a serious limitation of the CASEP approach. -The case base maintenance applied in CASEP has obtained a limited efficiency [73]. Experiments conducted on real data show that the proposed policy allows to reduce the case base size by at most 15% of its original size.…”

Section: Introductionmentioning

confidence: 99%

“…In order to improve CASEP system, we propose a new hybrid neuro-CBR system called CASEP2 [73,74], in which case length is variable and cases are indexed using an original temporal neural network [67] based on the principle of self-organizing maps [36,40]. Description and evaluation of this new system proposition are the main subject of this paper.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Intelligent GPS-based predictive engine control for a motor vehicle

Lee

Begg

Walters

et al. 2010

HIS

View full text Add to dashboard Cite

An intelligent Global Positioning System (GPS) based control system utilises information about the current vehicle position and upcoming terrain in order to reduce vehicle fuel consumption as well as improve road safety and comfort. The development of such in-vehicle control systems has provided static and dynamic road information. The vehicle running parameters have been mathematically defined whilst the engine control algorithms were derived from a custom-built engine test-rig. As the vehicle travelled over a particular route, road information such as gradient and position was stored with the past trajectory using a Neuro-Fuzzy technique. This road information was continuously updated and replaced by new data as the vehicle moved along, thereby adjusting the engine control parameters to reflect the actual current vehicle running data. The control system essentially used a fuzzy logic derived relief map of the test route and this was further validated and corrected based on the past trajectory from the in-vehicle GPS sensor. The simulation model demonstrated the feasibility and robustness of the control system for motor vehicle control applications

show abstract

Section: Applying Neural Network To User Behavior Modelingmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Intelligent GPS-based predictive engine control for a motor vehicle

Lee

Begg

Walters

et al. 2010

HIS

View full text Add to dashboard Cite

show abstract

“…Although, it is a promising data mining technique, CBR was rarely used for time series forecasting. Probably the main reason for this is that the use of case-based reasoning for time series processing introduces some new aspects that don't exist in the processing of "attribute-value" data [Zehraoui et al 2004]. These aspects concern mainly the case representation -the sequences can be very long and with different lengths.…”

Section: Case-based Reasoningmentioning

confidence: 99%

“…Several works have used the representation of a case with the whole sequence: a web navigation advisor -Radix system [Corvaisier et al 1997], the recommender system PADIM [Fuchs et al 1995] etc. The representation of a case with the subsequence was used in the recommendation approach BROADWAY [Jaczynski and Trousse 1998], the CASEP system [Zehraoui et al 2004] used for sequence prediction and the COBRA system [Malek and Kanawati 2001] that predicts the user actions in a web site. The representation of a case with a whole sequence allows avoiding an additional storage of more precise knowledge.…”

Section: Case-based Reasoningmentioning

confidence: 99%