Sharif Digital Repository

We study the problem of fair allocation for indivisible goods. We use the maxmin share paradigm introduced by Budish [16] as a measure for fairness. Kurokawa, Procaccia, and Wang [36] were the first to investigate this fundamental problem in the additive setting. They show that a maxmin guarantee (1-MMS allocation) is not always possible even when the number of agents is limited to 3. While the existence of an approximation solution (e.g. a 1/2-MMS allocation) is quite straightforward, improving the guarantee becomes subtler for larger constants. Kurokawa et al. [36] provide a proof for the existence of a 2/3-MMS allocation and leave the question open for better guarantees. Our main... 

Neural networks have been applied very successfully in the identification and control of dynamic systems. The universal approximation capabilities of the multilayer perceptron have made it a popular choice for modeling nonlinear systems and for implementing general-purpose nonlinear controllers. In this paper we try to model and control the mass-spring-damper mechanism as a 1 DOF system using neural networks. The control architecture used in this paper is Model reference controller (MRC) as one of the popular neural network control architectures  

We study the problem of maximizing constrained non-monotone submodular functions and provide approximation algorithms that improve existing algorithms in terms of either the approximation factor or simplicity. Different constraints that we study are exact cardinality and multiple knapsack constraints for which we achieve (0.25-)-factor algorithms. We also show, as our main contribution, how to use the continuous greedy process for non-monotone functions and, as a result, obtain a 0.13-factor approximation algorithm for maximization over any solvable down-monotone polytope  

In this paper, the design of an adaptive tracking control for a class of switched uncertain multiple-input–multiple-output nonlinear systems in the strict-feedback form with unmodeled dynamics in the presence of three types of input nonlinearity under arbitrary switching has been addressed. By means of an intelligent approximator like a fuzzy logic system or a neural network, the unknown dynamics are estimated. The unmodeled dynamics have been tackled with a dynamic signal. A universal framework for describing different types of input nonlinearity including saturation, backlash, and dead zone has been utilized. By applying the backstepping approach and the common Lyapunov function method,... 

The approximation power of general feedforward neural networks with piecewise linear activation functions is investigated. First, lower bounds on the size of a network are established in terms of the approximation error and network depth and width. These bounds improve upon state- of-the-art bounds for certain classes of functions, such as strongly convex functions. Second, an upper bound is established on the difference of two neural networks with identical weights but different activation functions. © The Author(s) 2018  

Wound healing is a complex biological process dependent on multiple variables: tissue oxygenation, wound size, contamination, etc. Many of these factors depend on multiple factors themselves. Mechanisms for some interactions between these factors are still unknown, presenting a barrier for scientists intending to model wound healing using an object-based programming approach. In this paper we focus on the neural networks and regard them as function approximators, and attempt to simulate remodeling phase of dermal wound healing process using neural networks as an intelligence technique. © 2009 IEEE  

This paper introduces an incremental local learning algorithm inspired by learning in neurobiological systems. This algorithm has no training phase and learns the world during operation, in a lifetime manner. It is a semi-supervised algorithm which combines soft competitive learning in input space and linear regression with recursive update in output space. This method is also robust to negative interference and compromises bias-variance dilemma. These qualities make the learning method a good nonlinear function approximator having possible applications in neuro-robotic systems. Some simulations illustrate the effectiveness of the proposed algorithm in function approximation, time-series... 

This paper deals with the boundary control problem for a certain class of linear infinite-dimensional systems commonly known as fractional-delay systems. It is assumed that the systems under consideration are, in general, described by multi-valued transfer functions. In this paper, we restrict our studies to a class of multi-valued transfer functions which are defined on a Riemann surface with limited number of Riemann sheets where the origin is a branch point. The proposed controller design algorithm is based on shaping the sensitivity function. Contrary to many other methods which approximate the underlying infinite-dimensional system by a finite-dimensional one and then apply a classical... 

This paper addresses the finite-time controller design for a class of nonlinear systems in the non-strict feedback form subject to unknown system dynamics and disturbances, arbitrary asymmetric time-varying output constraints, four types of input nonlinearities, and unknown control direction. Utilizing the barrier Lyapunov function (BLF) and backstepping technique, an adaptive finite-time controller has been proposed. The difficulties associating with non-strict feedback systems have been handled using the variable separation approach. Furthermore, the unknown control direction problem has been tackled by using the Nussbaum gain function. A unified framework has been utilized for handling... 

We analyze the interaction of a propagating guided electromagnetic wave with a quantum well embedded in a dielectric slab waveguide. First, we design a quantum well based on InAlGaAs compounds with the transition energy of 0.8 eV corresponding to a wavelength of 1.55 μm. By exploiting the envelope function approximation, we derive the eigenstates of electrons and holes and the transition dipole moments. Next, we calculate the electrical susceptibility of a three-level quantum system (as a model for the 2-D electron gas trapped in the waveguide), by using phenomenological optical Bloch equations. We show that the 2-D electron gas behaves as a conducting interface, whose conductivity can be... 

Residual stresses, created in a steel beam by elastic-plastic bending, are predicted using an approximate analysis and the finite element method. The predictions are compared to experimental measurements obtained from the application of incremental centre hole drilling, deep hole drilling and neutron diffraction methods. Finite element simulations of the incremental centre hole drilling and deep hole drilling methods applied to the predicted residual stresses permitted an assessment of their ability to reconstruct the stresses. An analytical reconstruction analysis using an Airy stress function together with boundary and equilibrium conditions is developed and applied to the predictions and... 

Wound healing is a complex biological process dependent on multiple variables: tissue oxygenation, wound size, contamination, etc. Many of these factors depend on multiple factors themselves. Mechanisms for some interactions between these factors are still unknown but it is generally accepted that collagen synthesis, accumulation and organization are increased by mechanical stimuli, resulting in a forced healing process which improves mechanical properties of the damaged tissue. In this paper we focus on the neural networks and regard them as function approximators, and attempt to simulate remodeling phase of dermal wound healing process using neural networks as an intelligent technique  

The activity coefficient at infinite dilution is a representative of the limiting non-ideality of a solute in a mixture. Various methods for the prediction of infinite dilution activity coefficients (IDACs) have been developed. Artificial neural networks are powerful mapping tools for nonlinear function approximations. Accordingly, an artificial neural network model is proposed for the prediction of the IDACs of binary systems where the properties of the individual components are used as inputs to the network. The input parameters of the neural network are the mixture temperature, critical temperature, critical pressure, critical volume, molecular weight, dipole moment and the acentric... 

Active Learning Method (ALM) is one of the powerful tools in soft computing and it is inspired by the human brain capabilities in approaching complicated problems. ALM, which is in essence an adaptive fuzzy learning algorithm, tries to model a Multi-Input Single-Output system with several single-input single-output subsystems. Each of these subsystems is then modeled by an ink drop spread (IDS) plane. IDS operator, which is the main processing engine of ALM, extracts two kinds of informative features, Narrow Path and Spread, from each IDS plane without complicated computations. These features from all IDS planes are then aggregated in the inference engine. Despite the great performance of... 

In this paper, we develop a stochastic programming model for the capacitated location-allocation problem in the heterogeneous environment where the demands are distributed according to the Bernoulli function with different probabilities. The capacitated sub-sources of facilities are also involved to satisfy customers’ demands in this work. This study aims to find optimal locations of facilities and optimal allocations of existing customers to the facilities so that the total cost of operating facilities, allocating the customers, expected servicing and outsourcing is minimized. Due to the large amount of customers with different demand probabilities, accurate estimation of the outsourcing... 

Active Learning Method (ALM) is one of the powerful tools in soft computing and it is inspired by the human brain capabilities in approaching complicated problems. ALM, which is in essence an adaptive fuzzy learning algorithm, tries to model a Multi-Input Single-Output system with several single-input single-output subsystems. Each of these subsystems is then modeled by an ink drop spread (IDS) plane. IDS operator, which is the main processing engine of ALM, extracts two kinds of informative features, Narrow Path and Spread, from each IDS plane without complicated computations. These features from all IDS planes are then aggregated in the inference engine. Despite the great performance of... 

In this paper, we present an envelope function analysis technique for the design of the emission spectra of a white quantum-well light-emitting diode (QWLED). The nano- metric heterostructure that we are dealing with is a multiple QW, consisting of periods of three single QWs with various well thicknesses. With the aid of 6 × 6 Luttinger Hamiltonian, we employ the combination of two methods, k · p perturbation and the transfer matrix method, to acquire the electron and hole wave functions numerically. The envelope function approximation was considered to obtain these wave functions for a special basis set. While adjacent valence sub-bands have been determined approximately, the conduction...