We prove adaptive bounds for learning algorithms that operate by making a sequence of choices. These adaptive bounds, which we call Microchoice bounds, can be used to make these algorithms self-bounding in the style o...
详细信息
ISBN:
(纸本)9781581131673
We prove adaptive bounds for learning algorithms that operate by making a sequence of choices. These adaptive bounds, which we call Microchoice bounds, can be used to make these algorithms self-bounding in the style of Freund [Fre98]. Furthermore, we can combine these bounds with Freund's more sophisticated query-tree approach, producing a modified query-tree structure that yields similar bounds to those in [Fre98] but that permits a much more efficient algorithmic approximation.
Network traffic classification has become vital due to the increased flow of web traffic from services like HTTP, FTP, and SMTP etc. The idea of Network traffic classification is to categorize the network traffic and ...
详细信息
We describe a framework for designing efficient active learning algorithms that are tolerant to random classification noise and differentially-private. The framework is based on active learning algorithms that are sta...
详细信息
We describe a framework for designing efficient active learning algorithms that are tolerant to random classification noise and differentially-private. The framework is based on active learning algorithms that are statistical in the sense that they rely on estimates of expectations of functions of filtered random examples. It builds on the powerful statistical query framework of Kearns [30]. We show that any efficient active statistical learning algorithm can be automatically converted to an efficient active learning algorithm which is tolerant to random classification noise as well as other forms of "uncorrelated" noise. We show that commonly studied concept classes including thresholds, rectangles, and linear separators can be efficiently actively learned in our framework. These results combined with our generic conversion lead to the first computationally-efficient algorithms for actively learning some of these concept classes in the presence of random classification noise that provide exponential improvement in the dependence on the error Ε over their passive counterparts. In addition, we show that our algorithms can be automatically converted to efficient active differentially-private algorithms. This leads to the first differentially-private active learning algorithms with exponential label savings over the passive case.
We introduce a new perception-based discriminative learning algorithm for labeling structured data such as sequences, trees, and graphs. Since it is fully kernelized and uses pointwise label prediction, large features...
详细信息
ISBN:
(纸本)1581138385
We introduce a new perception-based discriminative learning algorithm for labeling structured data such as sequences, trees, and graphs. Since it is fully kernelized and uses pointwise label prediction, large features, including arbitrary number of hidden variables, can be incorporated with polynomial time complexity. This is in contrast to existing labelers that can handle only features of a small number of hidden variables, such as Maximum Entropy Markov Models and Conditional Random Fields. We also introduce several kernel functions for labeling sequences, trees, and graphs and efficient algorithms for them.
In many programming educations, as a means of checking learner's comprehension relatively easily than code description, an assignment in the form of closed-ended questions is presented. The closed-ended question i...
详细信息
Despite the fact that different techniques have been developed to filter spam, due to the spammer’s rapid adoption of new spam detection techniques, we are still overwhelmed with spam emails. Currently, machine learn...
详细信息
This paper investigates the relationship between governance structure and bank performance in normal and crisis times. Using statistical learning algorithms on R, we regressed the profitability indicators as dependent...
详细信息
Human activity recognition is an extensively researched topic in the field of computer vision. Recognizing human activities without revealing a person's identity is one such use case. To solve this, we propose a p...
详细信息
We offer a new understanding of some aspects of practical SAT-solvers that are based on DPLL with unit-clause propagation, clause-learning, and restarts. On the theoretical side, we do so by analyzing a concrete algor...
详细信息
We present an approach to modeling the average case behavior of learning algorithms. Our motivation is to predict the expected accuracy of learning algorithms as a function of the number of training examples. We apply...
详细信息
暂无评论