???. Active Learning for Hidden Markov Models. Brigham Anderson, Andrew Moore brigham@cmu.edu , awm@cs.cmu.edu Computer Science Carnegie Mellon University. Outline. Active Learning Hidden Markov Models Active Learning + Hidden Markov Models. Notation. We Have: Dataset, D
An Image/Link below is provided (as is) to download presentation
Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.Content is provided to you AS IS for your information and personal use only.Download presentation by click this link.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.During download, if you can't get a presentation, the file might be deleted by the publisher.
E N D
More Related Content
Presentation Transcript
??? Active Learning for Hidden Markov Models Brigham Anderson, Andrew Moore brigham@cmu.edu, awm@cs.cmu.edu Computer Science Carnegie Mellon University
Outline • Active Learning • Hidden Markov Models • Active Learning + Hidden Markov Models
Notation • We Have: • Dataset, D • Model parameter space, W • Query algorithm, q
Uncertainty Sampling Random Sampling Uncertainty Sampling Random Sampling Uncertainty Sampling Random Sampling Uncertainty Sampling Random Sampling
Uncertainty Sampling GOOD: couldn’t be easier GOOD: often performs pretty well BAD: H(St) measures information gain about the samples, not the model Sensitive to noisy samples
Strategy #2:Query by Committee Temporary Assumptions: Pool Sequential P(W | D) Version Space Probabilistic Noiseless • QBC attacks the size of the “Version space”
O1 O2 O3 O4 O5 O6 O7 S1 S2 S3 S4 S5 S6 S7 FALSE! FALSE! Model #1 Model #2
O1 O2 O3 O4 O5 O6 O7 S1 S2 S3 S4 S5 S6 S7 TRUE! TRUE! Model #1 Model #2
O1 O2 O3 O4 O5 O6 O7 S1 S2 S3 S4 S5 S6 S7 FALSE! TRUE! Ooh, now we’re going to learn something for sure! One of them is definitely wrong. Model #1 Model #2
The Original QBCAlgorithm As each example arrives… • Choose a committee, C, (usually of size 2) randomly from Version Space • Have each member of C classify it • If the committee disagrees, select it.
Remember our whiny objection to Uncertainty Sampling? “H(St) measures information gain about the samples not the model.” BUT… If the source of the sample uncertainty is model uncertainty, then they equivalent! Why? Symmetry of mutual information.
Note: Pool-based again Note: No more Version space Dagan-Engelson QBC For each example… • Choose a committee, C, (usually of size 2) randomly from P(W | D) • Have each member C classify it • Compute the “Vote Entropy” to measure disagreement
How to Generate the Committee? • This important point is not covered in the talk. • Vague Suggestions: • Good conjugate priors for parameters • Importance sampling
Model Entropy P(W|D) P(W|D) P(W|D) W W W H(W) = high …better… H(W) = 0
Current model space entropy Expected model space entropy if we learn St Information-Gain • Choose the example that is expected to most reduce H(W) • I.e., Maximize H(W) – H(W | St)