Ryo Yoshinaka

Distribution Loss Minimization With Guaranteed Error Bound

Determining loss minimum configuration in a distribution network is a hard discrete optimization problem involving many variables. Since more and more dispersed generators are installed on the demand side of power systems and they are reconfigured frequently, developing automatic approaches is indispensable for effectively managing a large-scale distribution network. Existing fast methods employ local updates that gradually improve the loss to solve such an optimization problem. However, they eventually get stuck at local minima, resulting in arbitrarily poor results. In contrast, this paper presents a novel optimization method that provides an error bound on the solution quality. Thus, the obtained solution quality can be evaluated in comparison to the global optimal solution. Instead of using local updates, we construct a highly compressed search space using a binary decision diagram and reduce the optimization problem to a shortest path-finding problem. Our method was shown to be not only accurate but also remarkably efficient; optimization of a large-scale model network with 468 switches was solved in three hours with 1.56% relative error bound.

Identification in the Limit of k,l-Substitutable Context-Free Languages

Recently Clark and Eyraud (2005, 2007) have shown that substitutable context-free languages are polynomial-time identifiable in the limit from positive data. Substitutability in context-free languages can be thought of as the analogue of reversibility in regular languages. While reversible languages admit a hierarchy, namely k-reversible regular languages for each nonnegative integer k, Clark and Eyraud targeted the subclass of context-free languages that corresponds to zero-reversible regular languages only. Following Clark and Eyrauds proposal, this paper introduces a hierarchy of substitutable context-free languages as the analogue of that of k-reversible regular languages and shows that each class in the hierarchy is also polynomial-time identifiable in the limit from positive data.

Efficient learning of multiple context-free languages with multidimensional substitutability from positive data

Recently Clark and Eyraud (2007) [10] have shown that substitutable context-free languages, which capture an aspect of natural language phenomena, are efficiently identifiable in the limit from positive data. Generalizing their work, this paper presents a polynomial-time learning algorithm for new subclasses of multiple context-free languages with variants of substitutability.

Learning mildly context-sensitive languages with multidimensional substitutability from positive data

Recently Clark and Eyraud (2007) have shown that substitutable context-free languages, which capture an aspect of natural language phenomena, are efficiently identifiable in the limit from positive data. Generalizing their work, this paper presents a polynomialtime learning algorithm for new subclasses of mildly context-sensitive languages with variants of substitutability.

Polynomial-time identification of multiple context-free languages from positive data and membership queries

This paper presents an efficient algorithm that identifies a rich subclass of multiple context-free languages in the limit from positive data and membership queries by observing where each tuple of strings may occur in sentences of the language of the learning target. Our technique is based on Clark et al.s work (ICGI 2008) on learning of a subclass of context-free languages. Our algorithm learns those context-free languages as well as many non-context-free languages.

Distributional learning of parallel multiple context-free grammars

Natural languages require grammars beyond context-free for their description. Here we extend a family of distributional learning algorithms for context-free grammars to the class of Parallel Multiple Context-Free Grammars (pmcfgs). These grammars have two additional operations beyond the simple context-free operation of concatenation: the ability to interleave strings of symbols, and the ability to copy or duplicate strings. This allows the grammars to generate some non-semilinear languages, which are outside the class of mildly context-sensitive grammars. These grammars, if augmented with a suitable feature mechanism, are capable of representing all of the syntactic phenomena that have been claimed to exist in natural language.We present a learning algorithm for a large subclass of these grammars, that includes all regular languages but not all context-free languages. This algorithm relies on a generalisation of the notion of distribution as a function from tuples of strings to entire sentences; we define nonterminals using finite sets of these functions. Our learning algorithm uses a nonprobabilistic learning paradigm which allows for membership queries as well as positive samples; it runs in polynomial time.

Finding All Solutions and Instances of Numberlink and Slitherlink by ZDDs

Link puzzles involve finding paths or a cycle in a grid that satisfy given local and global properties. This paper proposes algorithms that enumerate solutions and instances of two link puzzles, Slitherlink and Numberlink, by zero-suppressed binary decision diagrams (ZDDs). A ZDD is a compact data structure for a family of sets provided with a rich family of set operations, by which, for example, one can easily extract a subfamily satisfying a desired property. Thanks to the nature of ZDDs, our algorithms offer a tool to assist users to design instances of those link puzzles.

The complexity and generative capacity of lexicalized abstract categorial grammars

Previous studies have shown that some well-known classes of grammars can be simulated by Abstract Categorial Grammars (de Groote 2001) in straightforward ways. These classes of grammars all generate subclasses of the PTIME languages. While the exact generative capacity of the class of ACGs and the complexity of its universal membership problem are both unknown, we show that the universal membership problem for the class of lexicalized ACGs is NP-complete and the languages generated by lexicalized ACGs form a subclass of NP which includes some NP-complete languages.

Counterexamples to the long-standing conjecture on the complexity of BDD binary operations

In this article, we disprove the long-standing conjecture, proposed by R.E. Bryant in 1986, that his binary decision diagram (BDD) algorithm computes any binary operation on two Boolean functions in linear time in the input-output sizes. We present Boolean functions for which the time required by Bryant@?s algorithm is a quadratic of the input-output sizes for all nontrivial binary operations, such as @?, @?, and @?. For the operations @? and @?, we show an even stronger counterexample where the output BDD size is constant, but the computation time is still a quadratic of the input BDD size. In addition, we present experimental results to support our theoretical observations.

Distributional learning of simple context-free tree grammars

This paper demonstrates how existing distributional learning techniques for context-free grammars can be adapted to simple context-free tree grammars in a straightforward manner once the necessary notions and properties for string languages have been redefined for trees. Distributional learning is based on the decomposition of an object into a substructure and the remaining structure, and on their interrelations. A corresponding learning algorithm can emulate those relations in order to determine a correct grammar for the target language.