What question did this study set out to answer?

This research investigates how class noise impacts supervised learning accuracy in medical applications and explores feature extraction as a solution.

January 1, 2006

Class Noise and Supervised Learning in Medical Domains: The Effect of Feature Extraction

View Full Paper

Ask AI

Bookmark

Authors

MPMykola PechenizkiyEindhoven University of Technology ATAlexey TsymbalUniversity College Dublin SPSeppo PuuronenTrinity College Dublin

Discussion

Member takes

Implication

Randomized trial analyzes class noise effects on supervised learning in medical domains, suggesting improved accuracy through feature extraction.

Key Points

This research investigates how class noise impacts supervised learning accuracy in medical applications and explores feature extraction as a solution.
Analyzed 8 medical datasets focusing on class noise and its effects.
Proposed feature extraction as a pre-processing method to reduce noise impact.
Evaluated classification accuracy of models using feature extraction compared to noisy data.
Feature extraction significantly improved classification accuracy on all tested datasets.
The approach did not require explicit elimination of noisy instances, simplifying data handling.
Achieved higher accuracy in learned models indicating the effectiveness of feature extraction.

Cite This Study

Pechenizkiy et al. (2006) studied this question.

synapsesocial.com/papers/6a1cc56c7108ba32b01d8fff https://doi.org/10.1109/cbms.2006.65

View Full Paper

Ask AI

Bookmark

Implication

Randomized trial analyzes class noise effects on supervised learning in medical domains, suggesting improved accuracy through feature extraction.

Class Noise and Supervised Learning in Medical Domains: The Effect of Feature Extraction

Discussion

Member takes

Implication

Key Points

Cite This Study

Class Noise and Supervised Learning in Medical Domains: The Effect of Feature Extraction

Discussion

Member takes

Implication

Key Points

Cite This Study

Also Consider

Also Consider