Stream mining project

A repo for data stream mining project (M2 D&K).

In this project, we implement the weighted ensemble classifier and its extension for cost-sensitive applications as described in [1].

Team members

• Armita KHAJEHNASSIRI
• Soumeya KAADA
• Minh Huong LE NGUYEN

Source code

• ensemble.py contains the implementation of Weighted Ensemble Classifier, as described in Algorithm 1 in [1].
• costsensiv_ensemble.py contains the implementation of Cost-Sensitive Weighted Ensemble Classifier, as described in Algorithm 2 and 3 in [1].

References

[1] "Mining Concept-Drifting Data Streams using Ensemble Classifiers", Haixun WANG, Wei FAN, Philip S. YU, Jiawei HAN.