Showing papers by "Amirkabir University of Technology published in 2018"

Adversarially Learned One-Class Classifier for Novelty Detection

Abstract: Novelty detection is the process of identifying the observation(s) that differ in some respect from the training observations (the target class). In reality, the novelty class is often absent during training, poorly sampled or not well defined. Therefore, one-class classifiers can efficiently model such problems. However, due to the unavailability of data from the novelty class, training an end-to-end deep network is a cumbersome task. In this paper, inspired by the success of generative adversarial networks for training deep models in unsupervised and semi-supervised settings, we propose an end-to-end architecture for one-class classification. Our architecture is composed of two deep networks, each of which trained by competing with each other while collaborating to understand the underlying concept in the target class, and then classify the testing samples. One network works as the novelty detector, while the other supports it by enhancing the inlier samples and distorting the outliers. The intuition is that the separability of the enhanced inliers and distorted outliers is much better than deciding on the original samples. The proposed framework applies to different related applications of anomaly and outlier detection in images and videos. The results on MNIST and Caltech-256 image datasets, along with the challenging UCSD Ped2 dataset for video anomaly detection illustrate that our proposed method learns the target class effectively and is superior to the baseline and state-of-the-art methods.

PLGA-Based Nanoparticles in Cancer Treatment.

Abstract: Nanomedicines can be used for a variety of cancer therapies including tumor-targeted drug delivery, hyperthermia, and photodynamic therapy. Poly (lactic-co-glycolic acid) (PLGA)-based materials are frequently used in such setups. This review article gives an overview of the properties of previously reported PLGA nanoparticles (NPs), their behavior in biological systems, and their use for cancer therapy. Strategies are emphasized to target PLGA NPs to the tumor site passively and actively. Furthermore, combination therapies are introduced that enhance the accumulation of NPs and, thereby, their therapeutic efficacy. In this context, the huge number of reports on PLGA NPs used as drug delivery systems in cancer treatment highlight the potential of PLGA NPs as drug carriers for cancer therapeutics and encourage further translational research.

A comprehensive review of heuristic optimization algorithms for optimal combined heat and power dispatch from economic and environmental perspectives

Abstract: Combined heat and power economic dispatch (CHPED) aims to minimize the operational cost of heat and power units satisfying several equality and inequality operational and power network constraints. The CHPED should be handled considering valve-point loading impact of the conventional thermal plants, power transmission losses of the system, generation capacity limits of the production units, and heat-power dependency constraints of the cogeneration units. Several conventional optimization algorithms have been firstly presented for providing the optimal production scheduling of power and heat generation units. Recently, experience-based algorithms, which are called heuristic and meta-heuristic optimization procedures, are introduced for solving the CHPED optimization problem. In this paper, a comprehensive review on application of heuristic optimization algorithms for the solution of CHPED problem is provided. In addition, the most popular heuristic and meta-heuristic optimization algorithms are studied in this paper, and implementation of the optimization procedures for the solution of CHPED problem taking into account the objective functions and different constrains are discussed. The main contributions of the reviewed papers are studied and discussed in details. Additionally, main considerations of equality and inequality constraints handled by different research studies are reported in this paper. Five test systems are considered for evaluating the performance of different optimization techniques. Optimal solutions obtained by employment of multiple heuristic and meta-heuristic optimization methods for test instances are demonstrated and the introduced methods are compared in terms of convergence speed, attained optimal solutions, and constrains. The best optimal solutions for five test systems are provided in terms of operational cost by employment of different optimization methods.

Classification of facility layout problems: a review study

Abstract: Facility layout problem (FLP) is defined as the placement of facilities in a plant area, with the aim of determining the most effective arrangement in accordance with some criteria or objectives under certain constraints, such as shape, size, orientation, and pick-up/drop-off point of the facilities. It has been over six decades since Koopmans and Beckmann published their seminal paper on modeling the FLP. Since then, there have been improvements to these researchers’ original quadratic assignment problem. However, research on many aspects of the FLP is still in its initial stage; hence, the issue is an interesting field to work on. Here, a review of literature is made by referring to numerous papers about FLPs. The study is mainly motivated by the current and prospective trends of research on such points as layout evolution, workshop characteristics, problem formulation, and solution methodologies. It points to gaps in the literature and suggests promising directions for future research on FLP.

A review on bio-fuel production from microalgal biomass by using pyrolysis method

Abstract: Due to the depletion of fossil fuels and their environmental issues, it is necessary to find energy resources which are renewable. Algal biomass becomes promising feedstock for bio-fuel production. They are considered as sustainable, renewable and effective biomass and bio-fuels obtained from them are more environment-friendly than fossil fuels. The aim of this work is to provide a state of the art review on pyrolysis of microalgae for generation of bio-fuels. Initially, some general aspects of biomass such as microalgae characteristics, different thermochemical processes and advantages of microalgae pyrolysis to produce bio-fuels are discussed. Then, different pyrolysis methods are explained and parameters affecting the process are addressed. Bio-fuels including gaseous, solid and liquid products have been characterized in a separate section. Finally, the technical challenges associated with microalgal pyrolysis commercialization are discussed in the last section of this article.

Adversarially Learned One-Class Classifier for Novelty Detection

Abstract: Novelty detection is the process of identifying the observation(s) that differ in some respect from the training observations (the target class). In reality, the novelty class is often absent during training, poorly sampled or not well defined. Therefore, one-class classifiers can efficiently model such problems. However, due to the unavailability of data from the novelty class, training an end-to-end deep network is a cumbersome task. In this paper, inspired by the success of generative adversarial networks for training deep models in unsupervised and semi-supervised settings, we propose an end-to-end architecture for one-class classification. Our architecture is composed of two deep networks, each of which trained by competing with each other while collaborating to understand the underlying concept in the target class, and then classify the testing samples. One network works as the novelty detector, while the other supports it by enhancing the inlier samples and distorting the outliers. The intuition is that the separability of the enhanced inliers and distorted outliers is much better than deciding on the original samples. The proposed framework applies to different related applications of anomaly and outlier detection in images and videos. The results on MNIST and Caltech-256 image datasets, along with the challenging UCSD Ped2 dataset for video anomaly detection illustrate that our proposed method learns the target class effectively and is superior to the baseline and state-of-the-art methods.

Rock strength estimation: a PSO-based BP approach

Abstract: Application of back-propagation (BP) artificial neural network (ANN) as an accurate, practical and quick tool in indirect estimation of uniaxial compressive strength (UCS) of rocks has recently been highlighted in the literature. This is mainly due to difficulty in direct determination of UCS in laboratory as preparing the core samples for this test is troublesome and time-consuming. However, ANN technique has some limitations such as getting trapped in local minima. These limitations can be minimized by combining the ANNs with robust optimization algorithms like particle swarm optimization (PSO). This paper gives insight into development of a hybrid PSO–BP predictive model of UCS. For this reason, dataset comprising the results of 228 laboratory tests including dry density, moisture content, P wave velocity, point load index test, slake durability index and UCS was prepared. These tests were conducted on 38 sandstone samples which were taken from two excavation sites in Malaysia. Findings showed that PSO–BP model performs well in predicting UCS. Nevertheless, to compare the prediction performance of the PSO–BP model, the UCS is predicted using ANN-based PSO and BP models. The correlation coefficient, R, values equal to 0.988 and 0.999 for training and testing datasets, respectively, suggest that the PSO–BP model outperforms the other predictive models.

Applications of Particle Swarm Optimization in Geotechnical Engineering: A Comprehensive Review

Abstract: Particle swarm optimization (PSO) is an evolutionary computation approach to solve nonlinear global optimization problems. The PSO idea was made based on simulation of a simplified social system, the graceful but unpredictable choreography of birds flock. This system is initialized with a population of random solutions that are updated during iterations. Over the last few years, PSO has been extensively applied in various geotechnical engineering aspects such as slope stability analysis, pile and foundation engineering, rock and soil mechanics, and tunneling and underground space design. A review on the literature shows that PSO has utilized more widely in geotechnical engineering compared with other civil engineering disciplines. This is due to comprehensive uncertainty and complexity of problems in geotechnical engineering which can be solved by using the PSO abilities in solving the complex and multi-dimensional problems. This paper provides a comprehensive review on the applicability, advantages and limitation of PSO in different disciplines of geotechnical engineering to provide an insight to an alternative and superior optimization method compared with the conventional optimization techniques for geotechnical engineers.

Nanophotonic Platforms for Enhanced Chiral Sensing

Abstract: Chirality plays an essential role in life, providing unique functionalities to a wide range of biomolecules, chemicals, and drugs, which makes chiral sensing and analysis critically important. The wider application of chiral sensing continues to be constrained by the involved chiral signals being inherently weak. To remedy this, plasmonic and dielectric nanostructures have recently been shown to offer a viable route for enhancing weak circular dichroism (CD) effects, but most relevant studies have thus far been ad hoc, not guided by a rigorous analytical methodology. Here, we report the first analytical treatment of CD enhancement and extraction from a chiral biolayer placed on top of a nanostructured substrate. We derive closed-form expressions of the CD and its functional dependence on the background-chiroptical response, substrate thickness and chirality, as well as on the optical chirality and intensity enhancement provided by the structure. Our results provide key insights into the trade-offs that ar...

Brain Tumor Classification via Convolutional Neural Network and Extreme Learning Machines

Abstract: Tumor identification is one of the main and most influential factors in the identification of the type of treatment, the treatment process, the success rate of treatment and the follow-up of the disease. Convolution neural networks are one of the most important and practical classes in the field of deep learning and feed-forward neural networks that is highly applicable for analyzing visual imagery. CNNs learn the features extracted by the convolution and maxpooling layers. Extreme Learning Machines (ELM) are a kind of learning algorithm that consists of one or more layers of hidden nodes. These networks are used in various fields such as classification and regression. By using a CNN, this paper tries to extract hidden features from images. Then a kernel ELM (KELM) classifies the images based on these extracted features. In this work, we will use a dataset to evaluate the effectiveness of our proposed method, which consists of three types of brain tumors including meningioma, glioma and pituitary tumor in T1-weighted contrast-enhanced MRI (CE-MRI) images. The results of this ensemble of CNN and KELM (KE-CNN) are compared with different classifiers such as Support Vector Machine, Radial Base Function, and some other classifiers. These comparisons show that the KE-CNN has promising results for brain tumor classification.

Optimizing an ANN model with ICA for estimating bearing capacity of driven pile in cohesionless soil

Abstract: The application of models provided by artificial neural network (ANN) in predicting bearing capacity of driven pile is underlined in several investigations. However, weakness of ANN in slow rate of convergence as well as finding reliable testing output is known to be the major drawbacks of implementing ANN-based techniques. The present study aims to introduce and evaluate an optimized ANN with imperialism competitive algorithm (ICA) model based to estimate bearing capacity of driven pile in cohesionless soil. The training data for optimizing the ICA-ANN structure are based on the in situ study. To develop the ICA-ANN model, the input parameters are internal friction angle of soil located in shaft (φ shaft), and tip (φ tip), pile length (L), effective vertical stress at pile toe (σ v), and pile area (A) while the output is the total driven pile bearing capacity in cohesionless soil. The predicted results are compared with a pre-developed ANN model to demonstrate the ability of the hybrid model. As a result, coefficient of determination (R 2) values of (0.885 and 0.894) and (0.964 and 0.974) was obtained for testing and training datasets of ANN and ICA-ANN models, respectively. In addition, values of variance account for (VAF) of (88.212 for training and 89.215 for testing) and (96.369 for training and 97.369 for testing, respectively) were obtained for ANN and ICA-ANN models, respectively. The obtained results declare high reliability of the developed ICA-ANN model. This model can be introduced as a new model in field of deep foundation engineering.

Nonlinear Load Sharing and Voltage Compensation of Microgrids Based on Harmonic Power-Flow Calculations Using Radial Basis Function Neural Networks

Abstract: A new decentralized hierarchical control scheme is presented to improve power sharing of multidistributed energy resources microgrids including nonlinear and sensitive loads. In this systems, electronically coupled distributed energy resources are responsible to perform the compensation to reduce the voltage harmonics at the point-of-common coupling. The proposed control scheme adds a new virtual impedance scheme, power calculation unit, and also a complementary loop to improve small- and large-signal stability margins and includes detailed modeling for all hierarchical control levels (either for grid-connected or islanded modes). Compared to conventional virtual impedance methods that add only line current feedforward terms to the voltage reference, here, the line current and voltage at the point-of-common coupling regulate the virtual impedance at fundamental and harmonic frequencies, respectively. So, mismatches in the feeder and line impedances are compensated. Moreover, a power calculation method based on harmonic power flow is presented, which exploits the nonlinear mapping ability of radial basis function neural networks to solve harmonic power flow and obtain voltage harmonics and active and reactive powers. To show the effectiveness of the proposed control scheme, offline time-domain simulation studies have been done on a test microgrid by MATLAB/SIMULINK software and verified experimentally using OPAL-RT real-time digital simulator.

An extended stepwise weight assessment ratio analysis (SWARA) method for improving criteria prioritization process

Abstract: The process of criteria prioritization and weighting is an important part of multiple attributes decision making. The most frequently applied multi-attribute decision-making weighting tools include analytical hierarchy process, stepwise weight assessment ratio analysis, factor relationship, and best---worst method. When policies are at the core of decision making, stepwise weight assessment ratio analysis method is the most efficient method for criteria evaluation. It involves two important steps: the first is to prioritize the criteria by consulting experts, while the second is the weighting process. This research seeks to extend stepwise weight assessment ratio analysis to improve the quality of the decision-making process by incorporating the reliability evaluation of experts' idea into the first step. Such a component is absent from the first step in all other similar models. Thus, an extended version of stepwise weight assessment ratio analysis can be applied for such evaluation. To test the applicability and performance of the proposed method, a numerical example from an earlier study was used. The proposed version can replace the classic version in future studies as an improved method in decision-making area.

Complete analysis and engineering applications of a megastable nonlinear oscillator

Abstract: In this paper, we investigate some unrevealed features of a newly introduced megastable chaotic oscillator. This oscillator has a rich dynamical behavior including limit cycle, torus and strange attractor. Also it has coexisting self excited and hidden attractors. Such multi-stable oscillator with coexisting self excited and hidden attractors is very rare in the literature. We have studied the oscillator dynamics using an analog circuit. Also a novel fuzzy-based robust and adaptive control method is designed to control this oscillator.