ebook img

Logistic regression for fast, accurate, and parameter - Komarix.org PDF

90 Pages·2008·2.19 MB·English
by  
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Logistic regression for fast, accurate, and parameter - Komarix.org

Logistic regression for fast, accurate, and parameter free data mining Paul Komarek Auton Lab Robotics Institute Carnegie Mellon University [email protected] http://www.komarix.org http://www.autonlab.org PaulKomarek,ConfidentialMaterials–p.1/44 Logistic Regression: Not Dead Yet Logistic regression (LR) is a venerable, but capable probabilistic binary classifier. LR is well-understood, mature and comfortable =) trusted LR accuracy is comparable to new-fangled state-of-the-art SVMs LR can be as fast or faster than linear SVMs Yet LR is often ignored in data mining literature. LR gets more press in text classification literature. Why do we care about LR speed? Why LR, instead of other binary classifiers? PaulKomarek,ConfidentialMaterials–p.2/44 LR is more useful when it is very fast 3 LR is useful already, even when estimation algorithms are O n . (cid:0) (cid:1) If LR algorithms were really fast, we could (obvious) do many logistic regressions (obvious) do many binary classifications which leads to . . . PaulKomarek,ConfidentialMaterials–p.3/44 Collaborative filtering – just a special multiclass problem, one LR model per “item” Link tasks Link completion (just collaborative filtering) TFF - Group detection MNOP - Alias detection AFDL - Links+Demographics Classification Many apps for fast binary classifiers Multiclass through voting (or an alternative) PaulKomarek,ConfidentialMaterials–p.4/44 Link tasks Link completion (just collaborative filtering) TFF - Group detection MNOP - Alias detection AFDL - Links+Demographics Classification Many apps for fast binary classifiers Multiclass through voting (or an alternative) Collaborative filtering – just a special multiclass problem, one LR model per “item” PaulKomarek,ConfidentialMaterials–p.4/44 Link completion (just collaborative filtering) TFF - Group detection MNOP - Alias detection AFDL - Links+Demographics Classification Many apps for fast binary classifiers Multiclass through voting (or an alternative) Collaborative filtering – just a special multiclass problem, one LR model per “item” Link tasks PaulKomarek,ConfidentialMaterials–p.4/44 TFF - Group detection MNOP - Alias detection AFDL - Links+Demographics Classification Many apps for fast binary classifiers Multiclass through voting (or an alternative) Collaborative filtering – just a special multiclass problem, one LR model per “item” Link tasks Link completion (just collaborative filtering) PaulKomarek,ConfidentialMaterials–p.4/44 MNOP - Alias detection AFDL - Links+Demographics Classification Many apps for fast binary classifiers Multiclass through voting (or an alternative) Collaborative filtering – just a special multiclass problem, one LR model per “item” Link tasks Link completion (just collaborative filtering) TFF - Group detection PaulKomarek,ConfidentialMaterials–p.4/44 AFDL - Links+Demographics Classification Many apps for fast binary classifiers Multiclass through voting (or an alternative) Collaborative filtering – just a special multiclass problem, one LR model per “item” Link tasks Link completion (just collaborative filtering) TFF - Group detection MNOP - Alias detection PaulKomarek,ConfidentialMaterials–p.4/44 Many apps for fast binary classifiers Multiclass through voting (or an alternative) Collaborative filtering – just a special multiclass problem, one LR model per “item” Link tasks Link completion (just collaborative filtering) TFF - Group detection MNOP - Alias detection AFDL - Links+Demographics Classification PaulKomarek,ConfidentialMaterials–p.4/44

Description:
Logistic regression (LR) is a venerable, but capable probabilistic binary classifier. .. regression may be better suited to rank-ordering predictions. Paul Komarek
See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.