Sharif Digital Repository

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... 

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... 

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... 

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... 

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... 

Until now many domain adaptation methods have been proposed. A major limitation of almost all of these methods is their assumption that all test data belong to a single stationary target distribution and a large amount of unlabeled data is available for modeling this target distribution. In fact, in many real world applications, such as classifying scene image with gradually changing lighting and spam email identification, data arrives sequentially and the data distribution is continuously evolving. In this thesis, we tackle the problem of adaptation to a continuously evolving target domain that has been recently introduced and propose the Evolving Domain Adaptation (EDA) method to classify... 

In multi-label classification problems each instance can simultaneously have multiple labels. In these problems, in addition to the complexities of the input feature space we encounter the complexities of output label space. In the multi-label classification problems, there are dependencies between different labels that need to be considered. Since the dimensionality of the label space in real-world applications can be (very) high, most methods which explicitly model these dependencies are ineffective in practice and recently those methods that transform the label space into a latent space have received attention. A class of these methods which uses output space dimension reduction, first... 

Due to the recent advances in the learning of multimodal data, humans tend to use computer systems in order to solve more complex problems. One of them is Visual Question Answering (VQA), where the goal is finding the answer of a question asked about the visual contents of a given image. This is an interdisciplinary problem between the areas of Computer Vision, Natural Language Processing and Reasoning. Because of recent achievements of Deep Neural Networks in these areas, recent works used them to address the VQA task. In this thesis, three different methods have been proposed which adding each of them to existing solutions to the VQA problem can improve their results. First method tries to... 

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... 

By the improvement of machine learning methods specially the Deep Learning in the last decade, there were expanding usage of these methods in Language Modeling task. As the essence of a language model is more basic, recently huge networks are trained with language model objective but fine-tuned on target tasks such as Question Answering, Sentiment Analysis and etc. which is a promising sign of its importance and usage in even other NLP tasks. However, this task still has severe problems. The Teacher Forcing based methods, suffer from the so-called exposure bias problem which is due to the train/test procedure discrepancy. Some solutions such as using Reinforcement Learning which has high... 

Liquid biopsy includes information about the progress of the tumor, the effectiveness of the treatment and the possibility of tumor metastasis. This type of biopsy obtains this information by doing diagnosis and enumerating genetic variations in cells and cell-free DNA (cfDNA). Only a small fraction of cfDNA which might be free circulation tumor DNA (ctDNA) fragments, has mutations and is usually identified by epigenetic variations. On the other hand, the use of liquid biopsy has decreased, and tumors in the final stages are often untreatable due to the low accuracy in prediction of cancer. In this research, the aim is to predict cancer using cfDNA methylation patterns. We obtain these... 

Image segmentation in medical imaging, as one of the most important branches of medical image analysis, often faces the challenge of limited labeled data for application in deep learning methods. The high cost of data collection and the need for expertise in image segmentation, particularly in three-dimensional images such as MRI and CT or sequence images like CMR, have all contributed to this problem, even for popular networks like U-Net, which struggle to achieve high accuracy. As a result, research efforts have focused on semi-supervised learning approaches, weakly supervised learning, as well as multi-instance learning in medical image segmentation. Unfortunately, each of these methods... 

A predefined assumption in many learning algorithms is that the training and test data must be in thesame feature space and have the same distribution.However, this assumption may not hold in all of these algorithms and in the real world there might be difference between the source and the targer domian, whether in the feature space or the distribution. Moreover, there might be a few number of labled data of the target domain which causes difficulty in learning an accurate classifier. In such cases, transferring knowledge can be useful if can be done successfully and transfer learning was introduced for this purpose. Domain Adaptation is one of the transfer leaning problems that assume some... 

Despite recent advancements in deep learning methods, these methods rely on a huge amount of training data to work. Recently the problem of solving classification and recently semantic segmentation problems with a few training data have gained attention to tackle this issue. In this research, we propose a meta-learning method by combining optimization-based and prototypical approaches in which a small portion of parameters are optimized with task-specific initialization. In addition to this and designing other parts of the method, we propose a new approach to use query data as an unlabeled sample to enhance task-specific learning. Alongside the mentioned method, we propose an approach to use... 

With the popularity of smart devices, a lot of applications have developed and deployed.Developers try to establish continuous interaction with their users by different tools including push notifications. Push notification is a message that is sent from developers to the users and as soon as the user’s device receives that, it appears on the device screen. Sending proper content to users in order to resume their engagement is one of the most important usages of notifications. Users are not interested in receiving irrelevant notifications, and receiving irrelevant notifications make them remove the application, so it’s important to predict users’ interest in different notifications and push... 

Solving the visual symbol grounding problem has long been a goal of modern aritificial intelligence. Due to recent breakthroughs in deep learning methods for natural language processing and visual interpretation tasks‚ the field now seems to be as near to achieving this goal as it ever was. Also recent progress in using recurrent neural netowrks (RNNs) for image description‚ has motivated the exploration of their application for video description tasks. However, while images remain static‚ interpreting videos require modeling complex dynamic temporal sturctures and then properly integrating that information into a natural language description. Recurrent neural networks can be both used to... 

Isoforms are mRNAs that are produced from a same gene site in the phenomenon called Alternative Splicing. Studies have shown that more than 95% of multiexon genes in humans have undergone Alternative Splicing. Although there are few changes in mRNA sequence, They may have a systematic effect on cell function and regulation. It is widely reported that isoforms of a gene have distinct or even contrasting functions. Most studies have shown that alternative splicing plays a significant role in human health and disease. Despite the wide range of gene function studies, there is little information about isoforms’ functionalities. Recently, some computational methods based on Multiple Instance... 

Graphs are widely used for representing structured data and analysis of them is an important area that appears in a broad domain of applications. Graph processing is of great importance in analyzing and predicting social media users' behavior, examining financial markets, detecting malware programs, and designing recombinant drugs. For example, consider a graph in which nodes and edges show the financial institutions and the financial connection between these institutions, respectively. Financial connection refers to the investment of one institute by another. Based on the graph structure, predicting trade stability and balance is extremely significant in macro decisions.In the last few... 

Brand perceptual mapping is a way to display the market structure in a particular industry. In the past, marketing managers relied on questionnaires and focus-groups' data to map the market structure and extract insights from it. These methods are expensive, time-consuming, and not generalizable due to the small sample size. With the expansion of social media, new techniques have been introduced that can map the market structure by analyzing user-generated content on social media. Currently, methods used for this task are automated methods based on the bag-of-word model, which ignores the meaning of words and sentences and needs big datasets. In this research, using language models and... 

Brand perceptual mapping is a way to display the market structure in a particular industry. In the past, marketing managers relied on questionnaires and focus-groups' data to map the market structure and extract insights from it. These methods are expensive, time-consuming, and not generalizable due to the small sample size. With the expansion of social media, new techniques have been introduced that can map the market structure by analyzing user-generated content on social media. Currently, methods used for this task are automated methods based on the bag-of-word model, which ignores the meaning of words and sentences and needs big datasets. In this research, using language models and...