We consider the problem of off-policy policy selection in reinforcement learning: using historical data generated from running one policy to compare two or more policies. We show that approaches based on importance sampling can be unfair—they can select the worse of two policies more often than not. We give two examples where the unfairness of importance sampling could be practically concerning...

The statistical method in a comparative study in which the standard treatment is theoretically or practically superior to the others has been investigated for a matched-pairs design. We derived this statistical method from one based on the maximum likelihood and score methods. Here we have shown that the score test statistic is algebraically the same as the statistic from the maximum likelihood...

Pure MLP-based approaches have not previously been t demonstrated to function well for continuous-speech recogni ion because of the need for accurate segmentation of the w speech signal. HMMs, on the other hand, provide a frame ork for simultaneous segmentation and classification of t speech, which has been demonstrated to be useful for con inuous recognition. Previous work by Morgan and Bourla...

− Low density parity-check codes are one of the hottest topics in coding theory nowadays. Equipped with very fast encoding and decoding algorithms (probabilistically,) LDPC are very attractive both theoretically and practically. In this paper, I present a throughout view of LDPC. I discuss in detail LDPC’s popular decoding algorithm: belief propagation. I also present some noticeable encoding a...

Evolutionary algorithms are randomized search strategies which have turned out to be efficient for optimization problems of quite different kind. In order to understand the behavior of evolutionary algorithms, one also is interested in examples where evolutionary algorithms need exponential time to find an optimal solution. Until now only artificial examples of this kind were known. Here an exa...

We present a new “cover and decomposition” attack on the elliptic curve discrete logarithm problem, that combines Weil descent and decomposition-based index calculus into a single discrete logarithm algorithm. This attack applies, at least theoretically, to all composite degree extension fields, and is particularly well-suited for curves defined over Fp6 . We give a real-size example of discret...

The long object problem is practically important and theoretically challenging. To solve the long object problem, spiral cone-beam CT was first proposed in 1991, and has been extensively studied since then. As a main feature of the next generation medical CT, spiral cone-beam CT has been greatly improved over the past several years, especially in terms of exact image reconstruction methods. Now...

ID: PSMG10043 P. Kuhlang1, S. Hempen2, W. Sihn1, J. Deuse2 1 Vienna University of Technology & Fraunhofer Austria, Theresianumgasse 27, 1004 Wien, Austria, 2 TU Dortmund University, Chair of Industrial Engineering, Dortmund, 44227, Germany Abstract A Process Management System in general represents is a suitable approach to improve processes in the broadest sense. For this improvement a lot of e...

Cryptography is to become familiar with the requirement of large, complex, information rich data sets for it’s privacy preservation. The privacy preserving data mining has been generated; to go through the concept of privacy in data mining is hard. Several algorithms and approaches are being generated theoretically, but practically it is hard. Privacy in data mining can be achieved through seve...

This paper demonstrates the analysis of the parameters of symmetrical full-wave dipole antenna based on multi – fed techniques. The current distributions were measured, followed by the measurement of parameters of antenna like gain, radiation patterns and input impedance. Based on these observations four equations for current distribution were eventually formulated. These equations were then us...