Adversarial Robustness of Deep Neural Networks in Text Domain

, M.Sc. Thesis Sharif University of Technology Behjati, Melika (Author) ; Soleymani Baghshah, Mahdieh (Supervisor)

Abstract

In recent years, neural networks have been widely used in most machine learning domains. However, it has been shown that these networks are vulnerable to adversarial examples. adversarial examples are small and imperceptible perturbations applied to the input which lead to producing wrong output and thus, fooling the network. This will become an important issue in security related applications of deep neural networks, such as self-driving cars and medical diagnostics. Since, in the wort-case scenario, even human lives could be threatened. Although, many works have focused on crafting adversarial examples for image data, only a few studies have been done on textual data due to the existing...

Test case prioritization using test case diversification and fault-proneness estimations

, Article Automated Software Engineering ; Volume 29, Issue 2 , 2022 ; 09288910 (ISSN) Mahdieh, M ; Mirian Hosseinabadi, S. H ; Mahdieh, M ; Sharif University of Technology

Springer 2022

Abstract

Regression testing activities greatly reduce the risk of faulty software release. However, the size of the test suites grows throughout the development process, resulting in time-consuming execution of the test suite and delayed feedback to the software development team. This has urged the need for approaches such as test case prioritization (TCP) and test-suite reduction to reach better results in case of limited resources. In this regard, proposing approaches that use auxiliary sources of data such as bug history can be interesting. We aim to propose an approach for TCP that takes into account test case coverage data, bug history, and test case diversification. To evaluate this approach we...

Microstructural characterisation of deformation behaviour of nickel base superalloy IN625

, Article Materials Science and Technology ; Volume 27, Issue 12 , 2011 , Pages 1858-1862 ; 02670836 (ISSN) Behjati, P ; Asgari, S ; Sharif University of Technology

2011

Abstract

Simple compression and microscopy techniques were employed to characterise the microstructural origin of the deformation behaviour of nickel base superalloy IN625 during large strain testing. The alloy exhibited a four-stage strain hardening response similar to that previously reported for low stacking fault energy face centred cubic alloys. At strains lower than about -0.06 (stage A), a falling regime of the hardening rate was observed. This stage was followed by a second stage (stage B) of slow increasing hardening rate, which was found to be coincident with the formation of Lomer-Cottrell locks. The second falling regime of strain hardening (stage C) was seen in the strain range of -0.25...

Biomechanical assessment of the niosh lifting equation in asymmetric load-handling activities using a detailed musculoskeletal model

, Article Human Factors ; Volume 61, Issue 2 , 2019 , Pages 191-202 ; 00187208 (ISSN) Behjati, M ; Arjmand, N ; Sharif University of Technology

SAGE Publications Inc 2019

Abstract

Objective: To assess adequacy of the National Institute for Occupational Safety and Health (NIOSH) Lifting Equation (NLE) in controlling lumbar spine loads below their recommended action limits during asymmetric load-handling activities using a detailed musculoskeletal model, that is, the AnyBody Modeling System. Background: The NIOSH committee employed simplistic biomechanical models for the calculation of the spine compressive loads with no estimates of the shear loads. It is therefore unknown whether the NLE would adequately control lumbar compression and shear loads below their recommended action limits during asymmetric load-handling activities. Method: Twenty-four static stoop lifting...

Improved K2 algorithm for Bayesian network structure learning

, Article Engineering Applications of Artificial Intelligence ; Volume 91 , 2020 Behjati, S ; Beigy, H ; Sharif University of Technology

Elsevier Ltd 2020

Abstract

In this paper, we study the problem of learning the structure of Bayesian networks from data, which takes a dataset and outputs a directed acyclic graph. This problem is known to be NP-hard. Almost most of the existing algorithms for structure learning can be classified into three categories: constraint-based, score-based, and hybrid methods. The K2 algorithm, as a score-based algorithm, takes a random order of variables as input and its efficiency is strongly dependent on this ordering. Incorrect order of variables can lead to learning an incorrect structure. Therefore, the main challenge of this algorithm is strongly dependency of output quality on the initial order of variables. The main...

Spectral Graph Partitioning

, M.Sc. Thesis Sharif University of Technology Behjati, Shahab (Author) ; Daneshgar, Amir (Supervisor)

Abstract

Graph partitioning, or graph clustering, is an essential researa problem in many areas. In this thesis, we focus on the partitioning problem of unweighted undirected graph, that is, graphs for which the weight of all edges is 1. We will investigate spectral methods for solving the graph partitioning and we compare them. In addition to theoretical analysis,We also implement some of spectral algorithms in matlab and apply them on standard graph data sets. Finally, the experimental
results obtained are offering

Event-Triggered Strategies for Model Predictive Control of Delayed Linear Systems

, M.Sc. Thesis Sharif University of Technology Behjati, Vahid (Author) ; Haeri, Mohammad (Supervisor)

Abstract

Delay is an important factor in modeling and describing many chemical processes, mechanical systems, teleoperation and networked control systems. On the other hand, model predictive control is a very useful and powerful tool for handling different types of systems including time delayed systems. The main disadvantage of model predictive control is massive computational load and which may lead to wasting system resources. One of the best options for dealing with this issue is to employ event-triggered strategy in model predictive control. Using this approach can relax computational efforts and save system resources. This thesis is focused on introducing a robust model predictive control using...

Fundamental Bounds for Clustering of Bernoulli Mixture Models

, M.Sc. Thesis Sharif University of Technology Behjati, Amin (Author) ; Motahari, Abolfazl (Supervisor)

Abstract

A random vector with binary components that are independent of each other is referred to as a Bernoulli random vector. A Bernoulli Mixture Model (BMM) is a combination of a finite number of Bernoulli models, where each sample is generated randomly according to one of these models. The important challenge is to estimate the parameters of a Bernoulli Mixture Model or to cluster samples based on their source models. This problem has applications in bioinformatics, image recognition, text classification, social networks, and more. For example, in bioinformatics, it pertains to clustering ethnic groups based on genetic data. Many studies have introduced algorithms for solving this problem without...

Evaluating the NIOSH Equation Performance to Estimate the Risk of Injury to Spine in Asymmetric Lifting

, M.Sc. Thesis Sharif University of Technology Behjati Ashtiani, Mohammad (Author) ; Arjmand, Navid (Supervisor)

Abstract

The 1991 NIOSH Lifting Equation (NLE) is widely used to assess the risk of injury to the spine by providing estimates of the recommended weight limit (RWL) in hands. The present study uses the AnyBody modeling software to verify whether the RWL generates L5-S1 within the limits (e.g., 3400 N for compression recommended by NIOSH and 1000 N for shear recommended in the literature). Twenty-four symmetric and asymmetric lifting activities were simulated to evaluate the RWL by the NLE and the L5-S1 loads by AnyBody. In two activities, involving large trunk flexion and 30 and 60 degrees of load asymmetry, the estimated RWL generated L5-S1 spine loads exceeding the recommended limits. The NIOSH...

Analysis and Simulation of a Graphene Based Plasmon Laser

, Ph.D. Dissertation Sharif University of Technology Behjati Ardakani, Sadreddin (Author) ; Faez, Rahim (Supervisor)

Abstract

In the present thesis, the SPASER is investigated, from theory to structure design. SPASER is a counterpart of laser in 3D subwavelength dimentions. In other words, SPASER does not suffer from the diffraction limit of photons which is a drawback in laser technology. It potentially can generate intense coherent dark and bright surface plasmon modes. The small size of SPASER gives it the capability of being integrated with electronic chips. So, the electronic technology will become faster if the SPASER is realized. In this thesis, we intend to use the unique plasmonic properties of graphene in our designs. Plasmons on graphene platforms have longer propagation length and larger lifetime in...

Failure analysis of holding U-bolts of an automobile wheels

, Article Engineering Failure Analysis ; Volume 16, Issue 5 , 2009 , Pages 1442-1447 ; 13506307 (ISSN) Behjati, P ; Etemadi, A. R ; Edris, H ; Sharif University of Technology

2009

Abstract

Holding U-bolts of an automobile wheels are manufactured from 10B21 boron steel rods. Changing the source of the rods from X to Y, with an identical manufacturing procedure, it was observed that parts fracture at the mechanical straightening stage (final stage of the manufacturing line). In this case study, microscopy techniques and mechanical tests were used to identify the cause of this failure. It was shown that higher contents of carbon and boron of the source Y rods promotes precipitation of boron containing carbides along the grain boundaries. These precipitates act as crack nucleation sites and are responsible for the fracture of the parts under the straightening stage stresses. Based...

Deep Zero-shot Learning

, M.Sc. Thesis Sharif University of Technology Shojaee, Mohsen (Author) ; Soleymani, Mahdieh (Supervisor)

Abstract

In some of object recognition problems, labeled data may not be available for all categories. Zero-shot learning utilizes auxiliary information (also called signatures) describing each category in order to find a classifier that can recognize samples from categories with no labeled instance. On the other hand, with recent advances made by deep neural networks in computer vision, a rich representation can be obtained from images that discriminates different categorizes and therefore obtaining a unsupervised information from images is made possible. However, in the previous works, little attention has been paid to using such unsupervised information for the task of zero-shot learning. In this...

Multi-Modal Distance Metric Learning

, M.Sc. Thesis Sharif University of Technology Roostaiyan, Mahdi (Author) ; Soleymani, Mahdieh (Supervisor)

Abstract

In many real-world applications, data contain multiple input channels (e.g., web pages include text, images and etc). In these cases, supervisory information may also be available in the form of distance constraints such as similar and dissimilar pairs from user feedbacks. Distance metric learning in these environments can be used for different goals such as retrieval and recommendation. In this research, we used from dual-wing harmoniums to combining text and image modals to a unified latent space when similar-dissimilar pairs are available. Euclidean distance of data represented in this latent space used as a distance metric. In this thesis, we extend the dual-wing harmoniums for...

Fabrication and Characterization of Thermoplastic Starch Based Nanocomposite for Bone Scaffold

, M.Sc. Thesis Sharif University of Technology Mahdieh, Zahra (Author) ; Bagheri, Reza (Supervisor)

Abstract

This project aimed to fabricate the bone scaffolds with applying thermoplastic starch-based nano-biocomposites. The starting materials for this scaffold are as follows: thermoplastic starch, ethylene vinyl alcohol as the polymer matrix and nanoforsterite as the ceramic reinforcing phase. Furthermore, vitamin E was used as antioxidant for preserving starch against thermo-mechanical degradations. Likewise, 3D pore structure was developed using azo-dicarbonamide and water in injection moulding process. With blending thermoplastic starch and ethylene vinyl alcohol, some thermoplastic starch’s properties including degradation rate and water absorption were modified. In addition, having...

Unsupervised Domain Adaptation via Representation Learning

, M.Sc. Thesis Sharif University of Technology Gheisary, Marzieh (Author) ; Soleymani, Mahdieh (Supervisor)

Abstract

The existing learning methods usually assume that training and test data follow the same distribution, while this is not always true. Thus, in many cases the performance of these learning methods on the test data will be severely degraded. We often have sufficient labeled training data from a source domain but wish to learn a classifier which performs well on a target domain with a different distribution and no labeled training data. In this thesis, we study the problem of unsupervised domain adaptation, where no labeled data in the target domain is available. We propose a framework which finds a new representation for both the source and the target domain in which the distance between these...

Deep Learning for Multimodal Data

, M.Sc. Thesis Sharif University of Technology Rastegar, Sarah (Author) ; Soleymani, Mahdieh (Supervisor)

Abstract

Recent advances in data recording has lead to different modalities like text, image, audio and video. Images are annotated and audio accompanies video. Because of distinct modality statistical properties, shallow methods have been unsuccessful in finding a shared representation which maintains the most information about different modalities. Recently, deep networks have been used for extracting high-level representations for multimodal data. In previous methods, for each modality, one modality-specific network was learned. Thus, high-level representations for different modalities were extracted. Since these high-level representations have less difference than raw modalities, a shared...

Deep Learning For Recommender Systems

, M.Sc. Thesis Sharif University of Technology Abbasi, Omid (Author) ; Soleimani, Mahdieh (Supervisor)

Abstract

Collaborative fltering (CF) is one of the best and widely employed approaches in Recommender systems (RS). This approach tries to fnd some latent features for users and items so it would predict user rates with these features. Early CF methods used matrix factorization to learn users and items latent features. But these methods face cold start as well as sparsity problem. Recent years methods employ side information along with rating matrix to learn users and items latent features. On the other hand, deep learning models show great potential for learning effective representations especially when auxiliary information is sparse. Due to this feature of deep learning, we use deep learning to...

Adversarial Networks for Sequence Generation

, M.Sc. Thesis Sharif University of Technology Montahaei, Ehsan (Author) ; Soleymani, Mahdieh (Supervisor)

Abstract

Lots of essential structures can be modeled as sequences and sequences can be utilized to model the structures like molecules, graphs and music notes. On the other hand, generating meaningful and new sequences is an important and practical problem in different applications. Natural language translation and drug discovery are examples of sequence generation problem. However, there are substantial challenges in sequence generation problem. Discrete spaces of the sequence and challenge of the proper objective function can be pointed out.On the other, the baseline methods suffer from issues like exposure bias between training and test time, and the ill-defined objective function. So, the...

Improving Sampling Efficiency of Probabilistic Graphical Models

, M.Sc. Thesis Sharif University of Technology Mahdieh, Mohsen (Author) ; Beigy, Hamid (Supervisor)

Abstract

Deep learning methods have become more popular in the past years. These methods use complex network architectures to model rich, hierarchical datasets. Although most of the research has been centered around Discriminative models, however, recently a lot of research is focused on Deep Generative Models. Two of the pioneering models in this field are Generative Adversarial Networks and Variational Auto-Encoders. In addition, knowing the structure of data helps models to search in a narrower hypothesis space. Most of the structure in datasets are models using Probabilistic Graphical Models. Using this structural information, one can achieve better parameter estimations. In the case of...

Improving Graph Construction for Semi-supervised Learning in Computer Vision Applications

, M.Sc. Thesis Sharif University of Technology Mahdieh, Mostafa (Author) ; Rabiee, Hamid Reza (Supervisor)

Abstract

Semi-supervised Learning (SSL) is an extremely useful approach in many applications where unlabeled data can be easily obtained. Graph based methods are among the most studied branches in SSL. Since neighborhood graph is a key component in these methods, we focus on methods of graph construction in this project. Graph construction methods based on Euclidean distance have the common problem of creating shortcut edges. Shortcut edges refer to the edges which connect two nearby points that are far apart on the manifold. Specifically, we show both in theory and practice that using geodesic distance for selecting and weighting edges results in more appropriate neighborhood graphs. We propose an...