Search for: conditional-distribution
Article Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 6 July 2015 through 10 July 2015 ; Volume 9134 , 2015 , Pages 143-154 ; 03029743 (ISSN) ; 9783662476710 (ISBN) ; Etesami, O ; Gohari, A ; Sharif University of Technology
Springer Verlag 2015
A Santha-Vazirani (SV) source is a sequence of random bits where the conditional distribution of each bit, given the previous bits, can be partially controlled by an adversary. Santha and Vazirani show that deterministic randomness extraction from these sources is impossible. In this paper, we study the generalization of SV sources for nonbinary sequences. We show that unlike the binary case, deterministic randomness extraction in the generalized case is sometimes possible. We present a necessary condition and a sufficient condition for the possibility of deterministic randomness extraction. These two conditions coincide in “non-degenerate” cases. Next, we turn to a distributed setting. In...
Article Computer Vision and Image Understanding ; Volume 162 , 2017 , Pages 116-134 ; 10773142 (ISSN) ; Jamzad, M ; Sharif University of Technology
Machine learning algorithms often suffer from good generalization in testing domains especially when the training (source) and test (target) domains do not have similar distributions. To address this problem, several domain adaptation techniques have been proposed to improve the performance of the learning algorithms when they face accuracy degradation caused by the domain shift problem. In this paper, we focus on the non-homogeneous distributed target domains and propose a new latent subdomain discovery model to divide the target domain into subdomains while adapting them. It is expected that applying adaptation on subdomains increase the rate of detection in comparing with the situation...
Article SIAM Journal on Computing ; Volume 46, Issue 1 , 2017 , Pages 1-36 ; 00975397 (ISSN) ; Etesami, O ; Gohari, A ; Sharif University of Technology
Society for Industrial and Applied Mathematics Publications 2017
A Santha-Vazirani (SV) source is a sequence of random bits where the conditional distribution of each bit, given the previous bits, can be partially controlled by an adversary. Santha and Vazirani show that deterministic randomness extraction from these sources is impossible. In this paper, we study the generalization of SV sources for nonbinary sequences. We show that unlike the binary setup of Santha and Vazirani, deterministic randomness extraction in the generalized case is sometimes possible. In particular, if the adversary has access to s "nondegenerate" dice that are c-sided and can choose one die to throw based on the previous realizations of the dice, then deterministic randomness...
Article Communications in Statistics: Simulation and Computation ; Volume 40, Issue 5 , Dec , 2011 , Pages 685-693 ; 03610918 (ISSN) ; Abbasi, B ; Mahlooji, H ; Sharif University of Technology
Motivated by numerous applications in Monte Carlo techniques and as of late, in deriving non dominated solutions in multi-objective optimization problems, this article addresses generating uniform random variables (λi, λi ≥ 0, i = 1,..., n) over a simplex in ℝ2 (n ≥ 2), i.e., Σi=1 n λi = 1. In this article, first, conditional distribution of λi where Σi=1 n λi = 1 is derived and then inverse method is applied to generate random variables
Article Engineering Applications of Artificial Intelligence ; Volume 58 , 2017 , Pages 157-170 ; 09521976 (ISSN) ; Soleymani Baghshah, M ; Sharif University of Technology
Elsevier Ltd 2017
Traditional learning algorithms cannot perform well in scenarios where training data (source domain data) that are used to learn the model have a different distribution with test data (target domain data). The domain adaptation that intends to compensate this problem is an important capability for an intelligent agent. This paper presents a domain adaptation method which learns to adapt the data distribution of the source domain to that of the target domain where no labeled data of the target domain is available (and just unlabeled data are available for the target domain). Our method jointly learns a low dimensional representation space and an adaptive classifier. In fact, we try to find a...
Critical voltage, thermal buckling and frequency characteristics of a thermally affected GPL reinforced composite microdisk covered with piezoelectric actuator, Article Mechanics Based Design of Structures and Machines ; 2020 ; Ghabussi, A ; Forooghi, A ; Shavalipour, A ; Habibi, M ; won Jung, D ; Safa, M ; Sharif University of Technology
Taylor and Francis Inc 2020
Due to the remarkable progress in the field of the manufacturing process, smart composites have become the desired target for high-tech engineering applications. Accordingly, for the first time, thermal buckling, critical voltage and vibration response of a thermally affected graphene nanoplatelet reinforced composite (GPLRC) microdisk in the thermal environment are explored with the aid of generalized differential quadrature method (GDQM). Also, the current microstructure is coupled with a piezoelectric actuator (PIAC). The extended form of Halpin-Tsai micromechanics is used to acquire the elasticity of the structure, whereas, the variation of thermal expansion, Poisson’s ratio, and density...
Article Soil Dynamics and Earthquake Engineering ; Volume 39 , August , 2012 , Pages 61-77 ; 02677261 (ISSN) ; Azarbakht, A ; Tabandeh, A ; Akbar Golafshani, A ; Sharif University of Technology
The exact and two approximate conditional spectra are compared in this manuscript as a target spectrum for the purpose of ground motion selection. The considered site is a real offshore site located at South Pars Gas Field in the Persian Gulf region. This case study site is influenced by four major seismic area sources in which the deaggregation results confirm that many comparable seismic scenarios can be taken into account. Therefore, an alternative to the conventional approximate conditional spectrum is proposed that has a small deviation from the exact solution. In addition, the use of different conditioning status of the probabilistic seismic hazard deaggregation (i.e., occurrence...
Investigating the effects of schedulability conditions on the power efficiency of task scheduling in an embedded system, Article ISORC 2010 - 2010 13th IEEE International Symposium on Object/Component/Service-Oriented Real-Time Distributed Computing, 5 May 2010 through 6 May 2010, Carmona, Sevilla ; Volume 1 , 2010 , Pages 102-106 ; 9780769540375 (ISBN) ; Miremadi, S. G ; Politecnica; IEEE ; Sharif University of Technology
Power consumption, performance and reliability are the most important parameters in modern safety-critical distributed real-time embedded systems. This paper evaluates and compares different schedulability conditions in faulttolerant Rate-Monotonic (RM) and Earliest-Deadline-First (EDF) algorithms, with respect to their power efficiency. The primary-backup scheme is used to implement fault tolerance in the algorithms. To evaluate the algorithms, a software tool is developed that can simulate an embedded system consisting of n processors and m periodic tasks. The results show that depending on the different schedulability conditions, the EDF algorithm implemented with the Best-Fit policy is...
Article Composite Structures ; Volume 93, Issue 12 , 2011 , Pages 3239-3251 ; 02638223 (ISSN) ; Hojjati, M. H ; Farrahi, G ; Alashti, R. A ; Sharif University of Technology
In this paper, distributions of stress and strain components of rotating disks with non-uniform thickness and material properties subjected to thermo-elastic loading under different boundary conditions are obtained by semi-exact methods of Liao's homotopy analysis method (HAM), Adomian's decomposition method and He's variational iteration method (VIM). The materials are assumed to be perfectly elastic and isotropic. A two dimensional plane stress analysis is used. The distribution of temperature over the disk radius is assumed to have power forms with the higher temperature at the outer surface. The results of the three methods are compared with those obtained by Runge-Kutta's numerical...