<div>Hi,</div><div><br></div>Have you seen <a href="http://imbalanced-learn.org">http://imbalanced-learn.org</a>?<div><br></div><div>Best,</div><div>Chris<br><br><div class="gmail_quote"><div dir="ltr">On Tue, 19 Jun 2018 17:53 S Hamidizade, <<a href="mailto:hamidizade.s@gmail.com">hamidizade.s@gmail.com</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"><div>Hi</div><div><br></div><div>I would appreciate if you could let me know what is the best way to categorize the approaches which have been developed to deal with imbalance class problem?</div><div><p><strong>This <a href="https://www.sciencedirect.com/science/article/pii/S0020025513005124" rel="nofollow noreferrer" target="_blank"><font color="#0066cc">article</font></a> categorizes them into:</strong></p><ol><li>Preprocessing: includes oversampling, undersampling and hybrid methods,</li><li>Cost-sensitive learning: includes direct methods and meta-learning which the latter further divides into thresholding and sampling,</li><li>Ensemble techniques: includes cost-sensitive ensembles and data preprocessing in conjunction with ensemble learning.</li></ol><p><strong>The <a href="https://dl.acm.org/citation.cfm?id=2907070" rel="nofollow noreferrer" target="_blank"><font color="#0066cc">second</font></a> classification:</strong></p><ol><li>Data Pre-processing: includes distribution change and weighting the data space. One-class learning is considered as distribution change.</li><li>Special-purpose Learning Methods</li><li>Prediction Post-processing: includes threshold method and cost-sensitive post-processing</li><li>Hybrid Methods:</li></ol><p><strong>The third <a href="https://link.springer.com/article/10.1007/s13748-016-0094-0" rel="nofollow noreferrer" target="_blank"><font color="#0066cc">article</font></a>:</strong></p><ol><li>Data-level methods</li><li>Algorithm-level methods</li><li>Hybrid methods</li></ol><p>The last classification also considers output adjustment as an independent approach.</p><p>Could you please let me know the class-weight in the sklearn's classifiers e.g., logistic regression is classified into which category? Is it true to say:</p><p>In case of the first categorization, it falls into cost-sensitive learning</p><p>In case of the second taxonomy, it would be classified into the third category i.e., cost-sensitive post-processing</p><p>In case of the third classification, it should fall into algorithm level <span></span></p><p>Best regards,</p></div></div>
_______________________________________________<br>
scikit-learn mailing list<br>
<a href="mailto:scikit-learn@python.org" target="_blank">scikit-learn@python.org</a><br>
<a href="https://mail.python.org/mailman/listinfo/scikit-learn" rel="noreferrer" target="_blank">https://mail.python.org/mailman/listinfo/scikit-learn</a><br>
</blockquote></div></div>