Timezone: »

Adaptive Regularization of Weight Vectors
Yacov Crammer · Alex Kulesza · Mark Dredze

Wed Dec 09 03:30 PM -- 03:31 PM (PST) @ None

We present AROW, a new online learning algorithm that combines several properties of successful : large margin training, confidence weighting, and the capacity to handle non-separable data. AROW performs adaptive regularization of the prediction function upon seeing each new instance, allowing it to perform especially well in the presence of label noise. We derive a mistake bound, similar in form to the second order perceptron bound, which does not assume separability. We also relate our algorithm to recent confidence-weighted online learning techniques and empirically show that AROW achieves state-of-the-art performance and notable robustness in the case of non-separable data.

Author Information

Yacov Crammer (Technion)
Alex Kulesza (Google)
Mark Dredze (Johns Hopkins)

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors