In this post I am showing my understanding about the paper Adaptive Regularization of Weight Vectors: http://papers.nips.cc/paper/3848-adaptive-regularization-of-weight-vectors.pdf The paper aims to address the negatives of a previous algorithm called confidence weighted (CW) learning by introducing the algorithm Adaptive Regularization Of Weights (AGOW). CW and AGOW are both online learning algorithms, meaning updates happen after observing each …
Continue reading “Adaptive Regularization of Weight Vectors — Mathematical derivation”