efficient convergence to a correct grammar. We discuss implications for learning from overt data only, as well as other learning issues. We argue that Optimality Theory promotes confluence of the demands of more effective learnability and deeper linguistic explanation.

Abstract. This paper addresses the problem of improving the accuracy of an hypothesis output by a learning algorithm in the distribution-free (PAC) learning model. A concept class is learnable (or strongly learnable) if, given access to a Source of examples of the unknown concept, the learner

procedure for a rate of convergence which succeeds after at most B(a)-many mind changes. This (B, L)-learnability provides quantitative information strictly in between a full rate of convergence (obtainable in general only from semi-constructive proofs) and a rate of metastability in the sense of Tao

LEARNABILITY. The mathematical theory of language learnability (also known as learnability theory, grammar induction, or grammatical inference) deals with idealized “learning procedures ” for acquiring grammars on the basis of exposure to evidence about languages. In one classic paradigm, presented

Abstract not found

This paper was selected by a process of anonymous peer reviewing for presentation at

Abstract: Short summary of most important research results that explain why the work was done, what was accomplished, and how it pushed scientific frontiers or advanced the field. This summary will be used for archival purposes and will be added to a searchable DoD database. First, we addressed the problem of detecting the period in which information diffusion burst occurs from a single observed diffusion sequence under the assumption that the delay of the information propagation over a social network follows the exponential distribution. To be more precise, we formulated the problem of detecting the change points and finding the values of the time delay parameter in the exponential distribution as an optimization problem of maximizing the likelihood of generating the observed diffusion sequence. We devised an efficient iterative search algorithm for the change point detection whose time complexity is almost linear to the number of data points. We tested the algorithm against the real Twitter data of the 2011 Tohoku earthquake and tsunami, and experimentally confirmed that the algorithm is much more efficient than the exhaustive naive search and is much more accurate than the simple greedy search. Second, we addressed the problem of how people make their own decisions based on their neighbors ’ opinions. The model best suited to discuss this problem is the voter model and several

presented by Schapire in his paper "The strength of weak learnability", and represents an improvement over his results. The analysis of our algorithm provides general upper bounds on the resources required for learning in Valiant's polynomial PAC learning framework, which are the best general

estimation and pruning techniques. We also present results of applying this algorithm to sales data obtained from a large retailing company, which shows the effectiveness of the algorithm.

. The purpose of this paper is to propose the framework as an evaluation technique for visual programming environments. We apply it to two commercially-available dataflow languages (with further examples from other systems) and conclude that it is effective and insightful; other HCI-based evaluation