hunch_net hunch_net-2008 hunch_net-2008-309 knowledge-graph by maker-knowledge-mining
Source: html
Introduction: Here are some papers from ICML 2008 that I found interesting. Risi Kondor and Karsten Borgwardt , The Skew Spectrum of Graphs . This paper is about a new family of functions on graphs which is invariant under node label permutation. They show that these quantities appear to yield good features for learning. Sanjoy Dasgupta and Daniel Hsu . Hierarchical sampling for active learning. This is the first published practical consistent active learning algorithm. The abstract is also pretty impressive. Lihong Li , Michael Littman , and Thomas Walsh Knows What It Knows: A Framework For Self-Aware Learning. This is an attempt to create learning algorithms that know when they err, (other work includes Vovk ). It’s not yet clear to me what the right model for feature-dependent confidence intervals is. Novi Quadrianto , Alex Smola , TIberio Caetano , and Quoc Viet Le Estimating Labels from Label Proportions . This is an example of learning in a speciali
sentIndex sentText sentNum sentScore
1 This paper is about a new family of functions on graphs which is invariant under node label permutation. [sent-3, score-0.695]
2 They show that these quantities appear to yield good features for learning. [sent-4, score-0.099]
3 This is the first published practical consistent active learning algorithm. [sent-7, score-0.21]
4 It’s not yet clear to me what the right model for feature-dependent confidence intervals is. [sent-11, score-0.108]
5 This is an example of learning in a specialization of the offline contextual bandit setting. [sent-13, score-0.593]
6 Learning should be used to solve the diversity problem, and doing it in an online bandit-like setting is quite natural. [sent-15, score-0.436]
7 I believe the setting can be generalized to a setting with features without too much work. [sent-16, score-0.533]
8 This paper doesn’t need an abstract given the title . [sent-18, score-0.249]
9 I hadn’t previously appreciated how well a random forest works in high dimensions. [sent-19, score-0.189]
10 A paper about an online contextual bandit setting specialized to a multiclass realizable yet otherwise adversarial setting that yields a practical algorithm. [sent-23, score-1.471]
11 I’d like to add that I thought the conference organization (and the colocation with COLT and UAI ) are particularly well done, the best I’ve seen. [sent-24, score-0.087]
12 The key seems to be tight integration of the colocating conference programs and hordes of local volunteers making sure everything is working. [sent-25, score-0.342]
13 I was also happy to see a 10 years award for best paper 10 years ago. [sent-26, score-0.412]
wordName wordTfidf (topN-words)
[('bandit', 0.231), ('setting', 0.217), ('graphs', 0.175), ('knows', 0.175), ('contextual', 0.158), ('abstract', 0.138), ('multiclass', 0.13), ('skew', 0.117), ('ambuj', 0.117), ('colocating', 0.117), ('hordes', 0.117), ('karampatziakis', 0.117), ('kleinberg', 0.117), ('quoc', 0.117), ('label', 0.115), ('practical', 0.112), ('paper', 0.111), ('online', 0.111), ('invariant', 0.108), ('volunteers', 0.108), ('diversity', 0.108), ('le', 0.108), ('intervals', 0.108), ('walsh', 0.108), ('years', 0.107), ('caruana', 0.102), ('specialization', 0.102), ('offline', 0.102), ('hadn', 0.102), ('littman', 0.102), ('features', 0.099), ('active', 0.098), ('spectrum', 0.097), ('realizable', 0.097), ('hierarchical', 0.097), ('smola', 0.097), ('err', 0.097), ('appreciated', 0.097), ('vovk', 0.097), ('node', 0.093), ('family', 0.093), ('thomas', 0.093), ('lihong', 0.093), ('nikos', 0.093), ('high', 0.092), ('estimating', 0.09), ('dimensions', 0.09), ('specialized', 0.087), ('happy', 0.087), ('colocation', 0.087)]
simIndex simValue blogId blogTitle
same-blog 1 1.0 309 hunch net-2008-07-10-Interesting papers, ICML 2008
Introduction: Here are some papers from ICML 2008 that I found interesting. Risi Kondor and Karsten Borgwardt , The Skew Spectrum of Graphs . This paper is about a new family of functions on graphs which is invariant under node label permutation. They show that these quantities appear to yield good features for learning. Sanjoy Dasgupta and Daniel Hsu . Hierarchical sampling for active learning. This is the first published practical consistent active learning algorithm. The abstract is also pretty impressive. Lihong Li , Michael Littman , and Thomas Walsh Knows What It Knows: A Framework For Self-Aware Learning. This is an attempt to create learning algorithms that know when they err, (other work includes Vovk ). It’s not yet clear to me what the right model for feature-dependent confidence intervals is. Novi Quadrianto , Alex Smola , TIberio Caetano , and Quoc Viet Le Estimating Labels from Label Proportions . This is an example of learning in a speciali
2 0.18021731 432 hunch net-2011-04-20-The End of the Beginning of Active Learning
Introduction: This post is by Daniel Hsu and John Langford. In selective sampling style active learning, a learning algorithm chooses which examples to label. We now have an active learning algorithm that is: Efficient in label complexity, unlabeled complexity, and computational complexity. Competitive with supervised learning anywhere that supervised learning works. Compatible with online learning, with any optimization-based learning algorithm, with any loss function, with offline testing, and even with changing learning algorithms. Empirically effective. The basic idea is to combine disagreement region-based sampling with importance weighting : an example is selected to be labeled with probability proportional to how useful it is for distinguishing among near-optimal classifiers, and labeled examples are importance-weighted by the inverse of these probabilities. The combination of these simple ideas removes the sampling bias problem that has plagued many previous he
3 0.17310824 361 hunch net-2009-06-24-Interesting papers at UAICMOLT 2009
Introduction: Here’s a list of papers that I found interesting at ICML / COLT / UAI in 2009. Elad Hazan and Comandur Seshadhri Efficient learning algorithms for changing environments at ICML. This paper shows how to adapt learning algorithms that compete with fixed predictors to compete with changing policies. The definition of regret they deal with seems particularly useful in many situation. Hal Daume , Unsupervised Search-based Structured Prediction at ICML. This paper shows a technique for reducing unsupervised learning to supervised learning which (a) make a fast unsupervised learning algorithm and (b) makes semisupervised learning both easy and highly effective. There were two papers with similar results on active learning in the KWIK framework for linear regression, both reducing the sample complexity to . One was Nicolo Cesa-Bianchi , Claudio Gentile , and Francesco Orabona Robust Bounds for Classification via Selective Sampling at ICML and the other was Thoma
4 0.16223332 400 hunch net-2010-06-13-The Good News on Exploration and Learning
Introduction: Consider the contextual bandit setting where, repeatedly: A context x is observed. An action a is taken given the context x . A reward r is observed, dependent on x and a . Where the goal of a learning agent is to find a policy for step 2 achieving a large expected reward. This setting is of obvious importance, because in the real world we typically make decisions based on some set of information and then get feedback only about the single action taken. It also fundamentally differs from supervised learning settings because knowing the value of one action is not equivalent to knowing the value of all actions. A decade ago the best machine learning techniques for this setting where implausibly inefficient. Dean Foster once told me he thought the area was a research sinkhole with little progress to be expected. Now we are on the verge of being able to routinely attack these problems, in almost exactly the same sense that we routinely attack bread and but
5 0.1605316 269 hunch net-2007-10-24-Contextual Bandits
Introduction: One of the fundamental underpinnings of the internet is advertising based content. This has become much more effective due to targeted advertising where ads are specifically matched to interests. Everyone is familiar with this, because everyone uses search engines and all search engines try to make money this way. The problem of matching ads to interests is a natural machine learning problem in some ways since there is much information in who clicks on what. A fundamental problem with this information is that it is not supervised—in particular a click-or-not on one ad doesn’t generally tell you if a different ad would have been clicked on. This implies we have a fundamental exploration problem. A standard mathematical setting for this situation is “ k -Armed Bandits”, often with various relevant embellishments. The k -Armed Bandit setting works on a round-by-round basis. On each round: A policy chooses arm a from 1 of k arms (i.e. 1 of k ads). The world reveals t
6 0.15633464 388 hunch net-2010-01-24-Specializations of the Master Problem
7 0.14758764 252 hunch net-2007-07-01-Watchword: Online Learning
8 0.14521825 324 hunch net-2008-11-09-A Healthy COLT
9 0.13950621 403 hunch net-2010-07-18-ICML & COLT 2010
10 0.13695061 8 hunch net-2005-02-01-NIPS: Online Bayes
11 0.13092586 360 hunch net-2009-06-15-In Active Learning, the question changes
12 0.12937094 289 hunch net-2008-02-17-The Meaning of Confidence
13 0.12804525 311 hunch net-2008-07-26-Compositional Machine Learning Algorithm Design
14 0.11869878 279 hunch net-2007-12-19-Cool and interesting things seen at NIPS
15 0.1175918 317 hunch net-2008-09-12-How do we get weak action dependence for learning with partial observations?
16 0.11575945 127 hunch net-2005-11-02-Progress in Active Learning
17 0.11434552 293 hunch net-2008-03-23-Interactive Machine Learning
18 0.11352672 213 hunch net-2006-10-08-Incompatibilities between classical confidence intervals and learning.
19 0.11211126 439 hunch net-2011-08-01-Interesting papers at COLT 2011
20 0.11166263 251 hunch net-2007-06-24-Interesting Papers at ICML 2007
topicId topicWeight
[(0, 0.231), (1, 0.025), (2, 0.001), (3, -0.089), (4, 0.174), (5, -0.036), (6, -0.04), (7, -0.093), (8, -0.07), (9, 0.072), (10, 0.253), (11, 0.028), (12, -0.069), (13, 0.075), (14, -0.031), (15, 0.054), (16, -0.04), (17, 0.032), (18, 0.046), (19, 0.036), (20, -0.0), (21, 0.01), (22, 0.082), (23, 0.043), (24, 0.052), (25, 0.037), (26, -0.104), (27, -0.041), (28, -0.044), (29, 0.054), (30, 0.046), (31, 0.029), (32, 0.038), (33, 0.05), (34, 0.04), (35, 0.041), (36, -0.063), (37, -0.004), (38, -0.077), (39, 0.009), (40, -0.039), (41, 0.092), (42, -0.005), (43, -0.06), (44, 0.03), (45, -0.143), (46, 0.056), (47, -0.019), (48, -0.017), (49, 0.008)]
simIndex simValue blogId blogTitle
same-blog 1 0.96121836 309 hunch net-2008-07-10-Interesting papers, ICML 2008
Introduction: Here are some papers from ICML 2008 that I found interesting. Risi Kondor and Karsten Borgwardt , The Skew Spectrum of Graphs . This paper is about a new family of functions on graphs which is invariant under node label permutation. They show that these quantities appear to yield good features for learning. Sanjoy Dasgupta and Daniel Hsu . Hierarchical sampling for active learning. This is the first published practical consistent active learning algorithm. The abstract is also pretty impressive. Lihong Li , Michael Littman , and Thomas Walsh Knows What It Knows: A Framework For Self-Aware Learning. This is an attempt to create learning algorithms that know when they err, (other work includes Vovk ). It’s not yet clear to me what the right model for feature-dependent confidence intervals is. Novi Quadrianto , Alex Smola , TIberio Caetano , and Quoc Viet Le Estimating Labels from Label Proportions . This is an example of learning in a speciali
2 0.61416596 361 hunch net-2009-06-24-Interesting papers at UAICMOLT 2009
Introduction: Here’s a list of papers that I found interesting at ICML / COLT / UAI in 2009. Elad Hazan and Comandur Seshadhri Efficient learning algorithms for changing environments at ICML. This paper shows how to adapt learning algorithms that compete with fixed predictors to compete with changing policies. The definition of regret they deal with seems particularly useful in many situation. Hal Daume , Unsupervised Search-based Structured Prediction at ICML. This paper shows a technique for reducing unsupervised learning to supervised learning which (a) make a fast unsupervised learning algorithm and (b) makes semisupervised learning both easy and highly effective. There were two papers with similar results on active learning in the KWIK framework for linear regression, both reducing the sample complexity to . One was Nicolo Cesa-Bianchi , Claudio Gentile , and Francesco Orabona Robust Bounds for Classification via Selective Sampling at ICML and the other was Thoma
3 0.59612864 388 hunch net-2010-01-24-Specializations of the Master Problem
Introduction: One thing which is clear on a little reflection is that there exists a single master learning problem capable of encoding essentially all learning problems. This problem is of course a very general sort of reinforcement learning where the world interacts with an agent as: The world announces an observation x . The agent makes a choice a . The world announces a reward r . The goal here is to maximize the sum of the rewards over the time of the agent. No particular structure relating x to a or a to r is implied by this setting so we do not know effective general algorithms for the agent. It’s very easy to prove lower bounds showing that an agent cannot hope to succeed here—just consider the case where actions are unrelated to rewards. Nevertheless, there is a real sense in which essentially all forms of life are agents operating in this setting, somehow succeeding. The gap between these observations drives research—How can we find tractable specializations of
4 0.57628864 251 hunch net-2007-06-24-Interesting Papers at ICML 2007
Introduction: Here are a few of the papers I enjoyed at ICML. Steffen Bickel , Michael Brüeckner, Tobias Scheffer , Discriminative Learning for Differing Training and Test Distributions There is a nice trick in this paper: they predict the probability that an unlabeled sample is in the training set vs. the test set, and then use this prediction to importance weight labeled samples in the training set. This paper uses a specific parametric model, but the approach is easily generalized. Steve Hanneke A Bound on the Label Complexity of Agnostic Active Learning This paper bounds the number of labels required by the A 2 algorithm for active learning in the agnostic case. Last year we figured out agnostic active learning was possible. This year, it’s quantified. Hopefull soon, it will be practical. Sylvian Gelly , David Silver Combining Online and Offline Knowledge in UCT . This paper is about techniques for improving MoGo with various sorts of learning. MoGo has a fair
5 0.57223541 279 hunch net-2007-12-19-Cool and interesting things seen at NIPS
Introduction: I learned a number of things at NIPS . The financial people were there in greater force than previously. Two Sigma sponsored NIPS while DRW Trading had a booth. The adversarial machine learning workshop had a number of talks about interesting applications where an adversary really is out to try and mess up your learning algorithm. This is very different from the situation we often think of where the world is oblivious to our learning. This may present new and convincing applications for the learning-against-an-adversary work common at COLT . There were several interesing papers. Sanjoy Dasgupta , Daniel Hsu , and Claire Monteleoni had a paper on General Agnostic Active Learning . The basic idea is that active learning can be done via reduction to a form of supervised learning problem. This is great, because we have many supervised learning algorithms from which the benefits of active learning may be derived. Joseph Bradley and Robert Schapire had a P
6 0.56429559 403 hunch net-2010-07-18-ICML & COLT 2010
7 0.54992819 317 hunch net-2008-09-12-How do we get weak action dependence for learning with partial observations?
8 0.54868883 432 hunch net-2011-04-20-The End of the Beginning of Active Learning
9 0.5361402 400 hunch net-2010-06-13-The Good News on Exploration and Learning
10 0.53510296 269 hunch net-2007-10-24-Contextual Bandits
11 0.53347516 360 hunch net-2009-06-15-In Active Learning, the question changes
12 0.53186661 385 hunch net-2009-12-27-Interesting things at NIPS 2009
13 0.53013098 311 hunch net-2008-07-26-Compositional Machine Learning Algorithm Design
14 0.52794701 439 hunch net-2011-08-01-Interesting papers at COLT 2011
15 0.52493912 8 hunch net-2005-02-01-NIPS: Online Bayes
16 0.51930857 252 hunch net-2007-07-01-Watchword: Online Learning
17 0.50712675 310 hunch net-2008-07-15-Interesting papers at COLT (and a bit of UAI & workshops)
18 0.5019415 127 hunch net-2005-11-02-Progress in Active Learning
19 0.4934417 248 hunch net-2007-06-19-How is Compressed Sensing going to change Machine Learning ?
20 0.48391783 186 hunch net-2006-06-24-Online convex optimization at COLT
topicId topicWeight
[(3, 0.036), (6, 0.165), (9, 0.054), (17, 0.016), (21, 0.025), (27, 0.231), (38, 0.069), (51, 0.034), (53, 0.067), (55, 0.115), (77, 0.036), (94, 0.018), (95, 0.041)]
simIndex simValue blogId blogTitle
same-blog 1 0.92761272 309 hunch net-2008-07-10-Interesting papers, ICML 2008
Introduction: Here are some papers from ICML 2008 that I found interesting. Risi Kondor and Karsten Borgwardt , The Skew Spectrum of Graphs . This paper is about a new family of functions on graphs which is invariant under node label permutation. They show that these quantities appear to yield good features for learning. Sanjoy Dasgupta and Daniel Hsu . Hierarchical sampling for active learning. This is the first published practical consistent active learning algorithm. The abstract is also pretty impressive. Lihong Li , Michael Littman , and Thomas Walsh Knows What It Knows: A Framework For Self-Aware Learning. This is an attempt to create learning algorithms that know when they err, (other work includes Vovk ). It’s not yet clear to me what the right model for feature-dependent confidence intervals is. Novi Quadrianto , Alex Smola , TIberio Caetano , and Quoc Viet Le Estimating Labels from Label Proportions . This is an example of learning in a speciali
2 0.90126687 275 hunch net-2007-11-29-The Netflix Crack
Introduction: A couple security researchers claim to have cracked the netflix dataset . The claims of success appear somewhat overstated to me, but the method of attack is valid and could plausibly be substantially improved so as to reveal the movie preferences of a small fraction of Netflix users. The basic idea is to use a heuristic similarity function between ratings in a public database (from IMDB) and an anonymized database (Netflix) to link ratings in the private database to public identities (in IMDB). They claim to have linked two of a few dozen IMDB users to anonymized netflix users. The claims seem a bit inflated to me, because (a) knowing the IMDB identity isn’t equivalent to knowing the person and (b) the claims of statistical significance are with respect to a model of the world they created (rather than one they created). Overall, this is another example showing that complete privacy is hard . It may be worth remembering that there are some substantial benefits from the Netf
3 0.82261175 403 hunch net-2010-07-18-ICML & COLT 2010
Introduction: The papers which interested me most at ICML and COLT 2010 were: Thomas Walsh , Kaushik Subramanian , Michael Littman and Carlos Diuk Generalizing Apprenticeship Learning across Hypothesis Classes . This paper formalizes and provides algorithms with guarantees for mixed-mode apprenticeship and traditional reinforcement learning algorithms, allowing RL algorithms that perform better than for either setting alone. István Szita and Csaba Szepesvári Model-based reinforcement learning with nearly tight exploration complexity bounds . This paper and another represent the frontier of best-known algorithm for Reinforcement Learning in a Markov Decision Process. James Martens Deep learning via Hessian-free optimization . About a new not-quite-online second order gradient algorithm for learning deep functional structures. Potentially this is very powerful because while people have often talked about end-to-end learning, it has rarely worked in practice. Chrisoph
4 0.81601971 194 hunch net-2006-07-11-New Models
Introduction: How should we, as researchers in machine learning, organize ourselves? The most immediate measurable objective of computer science research is publishing a paper. The most difficult aspect of publishing a paper is having reviewers accept and recommend it for publication. The simplest mechanism for doing this is to show theoretical progress on some standard, well-known easily understood problem. In doing this, we often fall into a local minima of the research process. The basic problem in machine learning is that it is very unclear that the mathematical model is the right one for the (or some) real problem. A good mathematical model in machine learning should have one fundamental trait: it should aid the design of effective learning algorithms. To date, our ability to solve interesting learning problems (speech recognition, machine translation, object recognition, etc…) remains limited (although improving), so the “rightness” of our models is in doubt. If our mathematical mod
5 0.81264949 406 hunch net-2010-08-22-KDD 2010
Introduction: There were several papers that seemed fairly interesting at KDD this year . The ones that caught my attention are: Xin Jin , Mingyang Zhang, Nan Zhang , and Gautam Das , Versatile Publishing For Privacy Preservation . This paper provides a conservative method for safely determining which data is publishable from any complete source of information (for example, a hospital) such that it does not violate privacy rules in a natural language. It is not differentially private, so no external sources of join information can exist. However, it is a mechanism for publishing data rather than (say) the output of a learning algorithm. Arik Friedman Assaf Schuster , Data Mining with Differential Privacy . This paper shows how to create effective differentially private decision trees. Progress in differentially private datamining is pretty impressive, as it was defined in 2006 . David Chan, Rong Ge, Ori Gershony, Tim Hesterberg , Diane Lambert , Evaluating Online Ad Camp
6 0.80683869 259 hunch net-2007-08-19-Choice of Metrics
7 0.80548751 225 hunch net-2007-01-02-Retrospective
8 0.80510646 220 hunch net-2006-11-27-Continuizing Solutions
9 0.80430508 230 hunch net-2007-02-02-Thoughts regarding “Is machine learning different from statistics?”
10 0.80380273 235 hunch net-2007-03-03-All Models of Learning have Flaws
11 0.80302697 378 hunch net-2009-11-15-The Other Online Learning
12 0.80219573 343 hunch net-2009-02-18-Decision by Vetocracy
13 0.7988137 466 hunch net-2012-06-05-ICML acceptance statistics
14 0.79873091 12 hunch net-2005-02-03-Learning Theory, by assumption
15 0.79689097 227 hunch net-2007-01-10-A Deep Belief Net Learning Problem
16 0.79649496 131 hunch net-2005-11-16-The Everything Ensemble Edge
17 0.79598099 183 hunch net-2006-06-14-Explorations of Exploration
18 0.79465514 437 hunch net-2011-07-10-ICML 2011 and the future
19 0.79464561 179 hunch net-2006-05-16-The value of the orthodox view of Boosting
20 0.79452103 484 hunch net-2013-06-16-Representative Reviewing