Sharif Digital Repository

Nowadays, the number of digital images has increased so that the management of this volume of data needs an efficient system for browsing, categorising and searching. Automatic image annotation is designed for assigning tags to images for more accurate retrieval. Non-negative matrix factorisation (NMF) is a traditional machine learning technique for decomposing a matrix into a set of basis and coefficients under the non-negative constraints. In this study, the authors propose a two-step algorithm for designing an automatic image annotation system that employs the NMF framework for its first step and a variant of K-nearest neighbourhood as its second step. In the first step, a new multimodal... 

In this paper, a new approach toward material detection and classification, based on the spectral analysis of millimeter-wave images, using machine learning technique is proposed. The focus of this paper is to detect concealed dangerous materials. It is shown that by using adequate number of training data captured from different materials of interest, the trained machine could detect concealed dangerous materials with an acceptable accuracy. The training phase is performed with materials of varying thickness, shape, background, covering layers and distance. The training data is collected with laboratory experiments in the frequency range of 27-31 GHz with 51 frequency samples. The results... 

Entanglement is one of the key resources of quantum information science which makes identification of entangled states essential to a wide range of quantum technologies and phenomena. This problem is however both computationally and experimentally challenging. Here we use autoencoder neural networks to find semi-optimal set of incomplete measurements that are most informative for the detection of entangled states. We show that it is possible to find high-performance entanglement detectors with as few as three measurements. Also, with the complete information of the state, we develop a neural network that can identify all two-qubits entangled states almost perfectly. This result paves the way... 

Optimal investment and operations of integrated local energy systems (ILESs) require medium to long-term prediction of energy consumption. To forecast load profiles, deep recurrent neural networks (DRNNs) are becoming increasingly useful due to their capability of learning uncertainty and high variability of load profiles. However, to explore and choose a DRNN model, out of conceivably numerous configurations, depends entirely on the performing task. In this regard, we tune and compare seven DRNN variants on the task of medium and long-term predictions for heating and electricity consumption. The ultimate DRNN model outperforms two state-of-the-art machine learning techniques, namely... 

To understand how aerial accidents occur, the installation of flight data recorders (FDR) and cockpit voice recorders (CVR) have become mandatory by law. However, such devices play a passive role and are used once an accident has occurred. Recent advances in machine learning techniques and their application in solving engineering problems are the keys to creating a more active role for both FDR as well as CVR. Here, we investigate a new approach to bringing intelligence to an FDR (called I-FDR). Through a continuous data-mining process, an I-FDR could bring better situational awareness to the flight crew. An I-FDR, similar to the FDR, records all pertinent flying parameters. In addition,... 

Real networks often exhibit nontrivial topological features that do not occur in random graphs. The need for synthesizing realistic networks has resulted in development of various network models. In this paper, we address the problem of selecting and calibrating the model that best fits a given target network. The existing model fitting approaches mostly suffer from sensitivity to network perturbations, lack of the parameter estimation component, dependency on the size of the networks, and low accuracy. To overcome these limitations, we considered a broad range of network features and employed machine learning techniques such as genetic algorithms, distance metric learning, nearest neighbor... 

Nowadays MapReduce and its open source implementation, Apache Hadoop, are the most widespread solutions for handling massive dataset on clusters of commodity hardware. At the expense of a somewhat reduced performance in comparison to HPC technologies, the MapReduce framework provides fault tolerance and automatic parallelization without any efforts by developers. Since in many cases Hadoop is adopted to support business critical activities, it is often important to predict with fair confidence the execution time of submitted jobs, for instance when SLAs are established with end-users. In this work, we propose and validate a hybrid approach exploiting both queuing networks and support vector... 

Characteristics and way of behavior of attacks and infiltrators on computer networks are usually very difficult and need an expert. In addition; the advancement of computer networks, the number of attacks and infiltrations is also increasing. In fact, the knowledge coming from expert will lose its value over time and must be updated and made available to the system and this makes the need for expert person always felt. In machine learning techniques, knowledge is extracted from the data itself which has diminished the role of the expert. Various methods used to detect intrusions, such as statistical models, safe system approach, neural networks, etc., all weaken the fact that it uses all the... 

To have a sustainable economy and environment, several countries have widely inclined to the utilization of non-fossil fuels like biomass fuels to produce heat and electricity. The advantage of employing biomass for combustion is emerging as a potential renewable energy, which is regarded as a cheap fuel. Chemical constituents or elements are essential properties in biomass applications, which would be costly and labor-intensive to experimentally estimate them. One of the criteria to evaluate the energy of biomass from an economic perspective is the higher heating value (HHV). In the present work, we have applied multilayer perceptron artificial neural network (MLP-ANN), least-squares...