Sharif Digital Repository

A Brain–Computer Interface is a system that provides a direct pathway for communication between a brain and a computer device by processing signals from sensors measuring brain activity (here Electroencephalography signals). Brain signals are known to be stochastic, non-stationary, non-linear and highly noisy, Therfore Brain–Computer Interface Systems rely on signal preprocessing, feature extraction and use of machine learning methods in order to detect mental state of Brain–Computer Interface user. Current approaches addressing the problem are mainly based on supervised learning methods. In this Thesis, first some of freely obtainable datasets with motor or motor-imagery paradigms are... 

In traditional machine learning approaches to classification, one uses only a labeled set to train the classifier. Labeled instances however are often difficult, expensive, or time consuming to obtain, as they require the efforts of experienced human annotators. Meanwhile unlabeled data may be relatively easy to collect, but there has been few ways to use them. Semi-supervised learning addresses this problem by using large amount of unlabeled data, together with the labeled data, to build better classifiers. Because semi-supervised learning requires less human effort and gives higher accuracy.Formally, this intuition corresponds to estimating a label function f on the graph so that it... 

Keywords or keyphrases are of importance as the smallest unit of representing the meaning of a text. Automated Keyword Extraction (AKE), as one of the natural language processing tasks is used in various applications such as searching, indexing and information retrieval. Keywords of scientific articles are basically specified manually by their authors, whereas most of the information available on the internet lack such keywords. In this research, we endeavor to automatically extract keywords of a set of Persian paper abstracts using an unsupervised machine learning method. The method used is to extract a set of candidate phrases from the text, and to cluster the document words to find a set... 

Automatic grammar induction is one of attractive research topics in natural language processing field. Automatic grammar induction methods can be categorized into three main groups of supervised, semi-supervised and unsupervised methods based on the type of training data that they need. Unsupervised methods are more difficult than two other. Data Oriented Parsing (DOP) is one of successful methods in unsupervised group. This method has been trained by some examples of language as same as child, then it parses new sentences based on its training knowledge. The aim of this project is finding and improving performance of UDOP method on Persian language as a Free Word Order language. Results of... 

Learning of shape classes from a set of given data is of special practical importance in computer vision. Large variations in the surrounding objects makes the problem of shape learning a very di°cult one. These di°culties may vary according to diferences in pose and part articulations of shapes. Despite all of the problems, large applications of this problem in areas such as tracking, object recognition and text recognition inspires many groups to work on it. In this thesis we aim to investigate a solution for this problem  

Identifying anomalous events is one of the vital topics in research as it often leads to the detection of actionable and critical information such as intrusions, faults, and system failures. With its importance, there has been a substantial body of work for network anomaly detection using supervised and unsupervised machine learning techniques with their own strengths and weaknesses. In this work, we take advantage of both worlds of unsupervised and supervised learning methods. The basic process model we present in this paper includes (i) clustering the training data set to create referential labels, (ii) building a supervised learning model with the automatically produced labels, and (iii)... 

In this study, we combined supervised and unsupervised machine learning algorithms, included the benefits of true identification of good borrowers and costs of false identification of bad borrowers, and then proposed a model for predicting the default of loan applicants with a profit-based approach. The results show that our proposed model has the best performance in profit measure in comparison with individual supervised models. In fact, we first divided the data into two train sets and one test set. We have constructed our model by training unsupervised models on the first train set and supervised models on the second train set. The results of implementing the model on the Australian and... 

In the stochastic analysis, the rough path is a generalization of the concept of smooth paths, which creates a theory for examining the control partial equations arising from irregular signals such as the Wiener process. This theory, which has found wide application in machine learning, time series analysis, and solving stochastic differential equations, is an approach by which the relationship between systems with extreme fluctuations and nonlinear systems is properly expressed.Rough path theory is also interpreted as a generalization of Taylor's expansion for smooth functions. In this research, we try to study the concept of rough path theory and signature and study its various... 

Deep reinforcement learning has achieved better cumulative rewards than humans in many environments like Atari. One drawback of these methods is their data inefficiency which makes training time-consuming, and in some cases having this amount of data is infeasible. Meta reinforcement learning can use past experiences to enable agents to adapt to new tasks faster and makes neural networks to train in a short amount of time.One of the methods in meta reinforcement learning is inferring tasks which helps exploitation policy to have good performance in new tasks. There’s a need to improve exploration policy as well as exploitation policy by gaining informative transitions about the new task.... 

In recent years, there have been so many studies in the field of machine learning, aiming to exploit the correlation among tasks. Among those, there are some types of tasks called primal and dual, which output of one is the input of the other. Dual learning is a method in which the dual and primal tasks are trained together. Many AI tasks emerge in dual form, e.g., English to Persian translation vs. Persian to English translation and image classification vs. image generation.Recently, several methods have been proposed to utilize the correlation between dual tasks. These methods can be divided into three groups of data-level, model-level, and inference-level dual learning. They have been... 

Visual Question Answering (VQA) deep-learning systems tend to capture superficial statistical correlations in the training data because of strong language priors and fail to generalize to test data with a significantly different question-answer(QA) distribution. To address this issue, we introduce a Visually Directed Question Encoder to replace the commonly used RNNs in base models. our method uses visual features alongside word embeddings of question words to encode each word. As a result, the model is forced to look at the visual information relevant to each word and it no longer produces answers based on just the question itself. We evaluate our approach on the VQA generalization task... 

Semantic classification of words based on unsupervised learning methods is a challenging issue in computational lexical semantics. The goal of this field of study is to recognize the words that are in the same semantic classes; i.e., can have the same set of arguments. Among all word categories, verb is known as one the most important and is assumed as the central part of the sentence in certain linguistic theories such as case grammar and dependency grammar. Based on Levin’s idea, diathesis alternations and the similarity between these alternations are the clues for the semantic classification of verbs. This idea is verified in languages such as English and German with promising results.... 

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... 

Improvement of automatic speech recognition systems and the growth of audio data (such as broadcast news, voice mail, telephony conversations and meetings) have attracted plenty of research interest in the field of speech summarization. The goal of this dissertation is to improve the performance of the speech summarization in the domain of multi-party conversations, specifically meetings. Most of the previous work in this field are inheritted from the text summarization counterpart, whithout paying much attention to the discourse specific information of the multi-party conversations. The main idea of this work is to use discourse information to improve the accuracy of extracted summaries in... 

The purpose of this research is to normalize non-standard persian texts. We proposed a method to transfigure the texts with any non-standard structure into a formal and standard form. One of the major complications of the text normalization is the large variety of non-standard structures, and the fact that these diversities could not be classified in one constructional pattern. Furthermore, the concept of text normalization, in different situations, has multiple different definitions, and any of this settings needs a distinct normalization method. Supervised learning methods are not suitable for normalization due to variety of both standard and non-standard texts as well as the absence of... 

Pattern recognition in data streams using bounded memory and bounded time is a difficult task. There are many techniques for recognizing patterns but when we talk about data streams these algorithms became useless since there are no enough memory to store all data. In data stream model the entire data is not available at any time and we don’t have enough time processing each data.
In this thesis we consider current methods for recognizing patterns from a data streams. The goal pattern in this study was the minimum total area of k convex polygons encloses all data  

Road accidents are currently one the main existing problems and a big challenge in Iran that is putting the lives of Iranian citizens in danger. Each accident is the result of a complex interplay between road users, vehicles, roads and environment. One of the main goals of accident data analysis is to identify and determine the main factors of a road accident. The dataset used here was obtained from the road traffic police and is stored in 3 different databases and corresponds to the accidents that happened between years 1390 and 1395 according to the Shamsi calendar. In this thesis, in order to deal with the inherent complexity and heterogeneity of the accident data, we will first introduce... 

Due to the overall progress in the processing of imaging tissue cells, the identification and diagnosis of complex diseases using machine learning methods has become very important. Recognizing cell characteristics such as size, shape, and chromatin design is essential in determining cell type, which can be achieved through learning methods such as deep network training. Finding the nucleus or cytoplasm of cells in medical images is a small but significant part of a long process of diagnosing and treating diseases. Today, artificial intelligence has rushed to the aid of experts in this field and has increased the speed and accuracy of experts in finding these cells and their nuclei. This... 

Today, the production of colloquial texts in messengers, search engines, and question and answer systems has increased significantly, while text documents in other fields have a formal tone and style. Thus, there is a need for a system to convert these texts from colloquial form to the formal style. Attention to this need in non-Persian languages has also been recently and seriously felt, but almost at the time of writing, an efficient system has not been offered, and this issue requires more work in Persian than in languages such as English. In general, transferring texts from one form to another falls into the category of natural language processing applications and is called "style...