Online PassiveAggressive Algorithms在线被动攻击算法.pptVIP

Learning to Align Polyphonic Music Shai Shalev-Shwartz Hebrew University, Jerusalem Joint work with Yoram Singer, Google Inc. Joseph Keshet, Hebrew University Motivation Symbolic Representation Acoustic Representation The Alignment Problem Setting The Alignment Problem Setting Goal: learn an alignment function Previous Work Dynamic Programming (rule based) Dannenberg 1984 Soulez et al. 2003 Orio Schwarz 2001 Generative Approaches Raphael 1999 Durey Clements 2001 Shalev-Shwartz et al. 2002 Our Solution Why Discriminative Learning? Outline of Solution Define a quantitative assessment of alignments Define a hypotheses class - what is the form of our alignment functions : Map all possible alignments into vectors in an abstract vector-space Find a projection in the vector-space which ranks alignments according to their quality Suggest a learning algorithm Assessing alignments Feature Functions for Alignment Feature Functions for Alignment Main Solution Principle Main Solution Principle (cont.) Main Solution Principle (cont.) Hypotheses class The form of our alignment functions: predict the alignment which attains the highest projection defines the direction of projection Learning algorithm Optimization Problem: Given a training set: Find: a projection and a maximal confidence scalar  such that the data is ranked correctly: Algorithmic aspects Iterative algorithm: Works on one alignment example at a time The algorithm works in polynomial time although the number of constraints is exponentially large Simple to implement Convergence: Converges to a high confidence solution #iterations depends on the best attainable confidence Generalization: The gap between test and train error decreases with the #examples. The gap is bounded above by Experimental Results Task: alignment of polyphonic piano music Dataset: 12 musical pieces where sound and MIDI were both recorded + other performances of th