Sharif Digital Repository

Boosting is a general method for improving the accuracy of any given learning algorithm. In this paper, we have proposed the boosting method for face recognition (FR) that improves the linear discriminant analysis (LDA)-based technique. The improvement is achieved by incorporating the regularized LDA (R-LDA) technique into the boosting framework. R-LDA is based on a new regularized Fisher's discriminant criterion, which is particularly robust against the small sample size problem compared to the traditional one used in LDA. The AdaBoost technique is utilized within this framework to generalize a set of simple FR subproblems and their corresponding LDA solutions and combines the results from... 

The effects of resonator topology on the phase noise of LC oscillators are studied in this paper. It is shown that there is a neglected factor that can considerably influence the phase noise behavior of the oscillator. We designate this factor as the Inductance Energy Factor (IEF), which directly depends on the topology of the resonator. It is shown that through proper design of the resonator for a better IEF, the oscillator phase noise can be improved. It is also shown that by modifying the resonator structure to improve IEF, the phase noise does not have to be constrained by the minimum realizable inductance and its maximum quality factor. Therefore, the power-phase noise trade-off remains... 

Meet in the middle (MITM) attack is one of the most important and applicable methods for cryptanalysis of block ciphers. In this paper, a more generalized method for MITM attack is considered. For this purpose, a notion, namely cut-set, is utilized by which several numbers of MITM attacks can be performed. However, manual investigation on these cases is time-consuming and sometimes not error-free. Therefore, a new search algorithm is also provided to obtain proper attacks in a timely manner. For examination, this new search algorithm, which could make an automated attack along with some certain ideas, is applied on HIGHT, Piccolo-128, CRAFT and AES-128 block ciphers. The least time... 

In this paper a general method for system level prediction of INL in pipeline analog to digital converters is presented. For each stage of the ADC, a new error model consisting of an input referred gain error and a nonlinear term is introduced. An analytic method to calculate INL from all error sources is presented. INL model for a switched-capacitor implementation is also presented. ©2006 IEEE  

A general method for deriving electromagnetic particle in cell (EMPIC) algorithms has been given by Eastwood [Comput. Phys. Commun. 64, 252 (1991)]. This method devises variation of the action-integral to find discrete governing equations. The most important advantage of this method is automatic inclusion of the time coordinate via the action integral into the computational domain. This inclusion is inevitable because electromagnetic algorithms are based on time evolution of the system from its initial state. The drawback of this method is that it is rather abstract. This causes obscurity of particle-mesh interactions and makes it hard to analyze physical treatments of the computational...