260 likes | 460 Views
Pranking with Ranking Koby Crammer and Yoram Singer. Lecture: Dudu Yanay. The Problem . Input: Each instance is associated with a rank or a rating, i.e. an integer from ‘1’ to ‘K’.
E N D
Pranking with RankingKoby Crammer and Yoram Singer Lecture: DuduYanay
The Problem • Input:Each instance is associated with a rank or a rating, i.e. an integer from ‘1’ to ‘K’. • Goal:To find a rank-prediction rule which assigns to each instance a rank which is as close as possible to the instance true rank. • Similar problems: • Classifications. • Regression.
Natural Setting For… • Information Retrieval. • Collaborative filtering:Predict a user’s rating on new items (books, movies etc) given the user’s past rating of similar items.
Possible Solutions • To cast a rating problem as a regression problem. • To reduce a total order into a set of preferences over pairs. • Time consuming since it might require to increase the sample size from to .
Lets try another approach… • Online Algorithm (Littlestone 1988): • Each can be computedin polynomial time. • If the problem is separable,after polynomial failures(no) the learner doesn’t makea mistake. Meaning: מורה לומד Animation from Nader Bshouty’s Course.
The PERCEPTRON algorithm Animation from Nader Bshouty’s Course.
The PERCEPTRON algorithm A slide from Nader Bshouty’s Course.
PRank algorithm - The model • Input: A sequence • . • Output: A ranking rule where: • . • . • . • Ranking loss after T rounds is: where is the TRUE rank of the instance in round ‘t’ and .
PRank algorithm - The update rule • Given an input instance-rank pair , if: • . • . • Lets represent the above inequalities by where The TRUE rank vector
PRank algorithm - The update rule • Given an input instance-rank pair , if . • So, let’s “move” the values of and towards each other: • . • , where the sum is only over the indices ‘r’ for which there was a prediction error, i.e., .
The update rule - Illustrasion Correct interval Predicted Rank 1 2 3 4 5
The PRank algorithm Building the TRUErank vector Checking which thresholdprediction is wrong Updating the hypothesis
PRank Analysis – Consistent Hypothesis • First, we need to show that the output hypothesis of Prank is acceptable. Meaning, if and is the final ranking rule then . • Proof – By induction:Since the initialization of the thresholds is such that , then it suffices to show that the claim hold inductively. • Lemma 1 (Order Preservation):Let and be the current ranking rule, where and let be an instance-rank pair fed to Prank on round ‘t’. Denote by and the resulting ranking after the update of Prank, then
Lemma 1 – Proof Correct interval Predicted Rank Option 1 1 2 3 4 5 6 Predicted Rank Correct interval Option 2 1 2 3 4 5
PRank Analysis – Mistake bound • Theorem 2:Let be an input sequence for PRank where . and . Denote by . Assume that there is a ranking rule with of a unit norm that classifies the entire sequence correctly with margin . . Then, the rank loss of the algorithm , is at the most .
Experiments • Comparison between: • Prank. • MultiClassPerceptron – MCP. • Widrow-Hoff (online regression) – WH. • Datasets: • Synthetic. • EachMovie.
Synthetic Dataset • Randomly generated points - uniformly at random. • Each point was assign a rank according to: • - noise. • Generated 100 sequences of instance-rank pairs, each of length 7000.
EachMovie Dataset • Collaborative filtering dataset. Contains ratings of movies provided by 61,265 people. • 6 possible rating: 0, 0.2, 0.4, 0.6, 0.8, 1. • Only people with atleast 100 rating whereconsidered. • Chose at random oneperson to be the TRUE rank and otherratings where used asfeatures(-0.5,-0.3,-0.1,0.1, 0.3, 0.5).
EachMovie Dataset – cont’ • Batch setting • Ran Prank over the training data as an online algorithm and used its last hypothesis to rank the unseen data.
משפט PERCEPTRON הוכחה
משפט PERCEPTRON הוכחה
משפט PERCEPTRON הוכחה