Fatos T. Yarman Vural

BAS: a perceptual shape descriptor based on the beam angle statistics

The proposed shape descriptor is based on the beams originated from a boundary point, which are defined as lines connecting that point with the rest of the points on the boundary. At each point, the angle between a pair of beams is calculated to extract the topological structure of the boundary. Then, a shape descriptor is defined by using the third-order statistics of all the beam angles in a set of neighborhood systems. It is shown that beam angle statistics (BAS) is invariant to translation, rotation, scale and is insensitive to distortions. Experiments are done on the dataset of MPEG 7 Core Experiments Shape-1. It is observed that BAS outperforms the MPEG 7 shape descriptors.

Sparse Attack Construction and State Estimation in the Smart Grid: Centralized and Distributed Models

New methods that exploit sparse structures arising in smart grid networks are proposed for the state estimation problem when data injection attacks are present. First, construction strategies for unobservable sparse data injection attacks on power grids are proposed for an attacker with access to all network information and nodes. Specifically, novel formulations for the optimization problem that provide a flexible design of the trade-off between performance and false alarm are proposed. In addition, the centralized case is extended to a distributed framework for both the estimation and attack problems. Different distributed scenarios are proposed depending on assumptions that lead to the spreading of the resources, network nodes and players. Consequently, for each of the presented frameworks a corresponding optimization problem is introduced jointly with an algorithm to solve it. The validity of the presented procedures in real settings is studied through extensive simulations in the IEEE test systems.

Machine Learning Methods for Attack Detection in the Smart Grid

Attack detection problems in the smart grid are posed as statistical learning problems for different attack scenarios in which the measurements are observed in batch or online settings. In this approach, machine learning algorithms are used to classify measurements as being either secure or attacked. An attack detection framework is provided to exploit any available prior knowledge about the system and surmount constraints arising from the sparse structure of the problem in the proposed approach. Well-known batch and online learning algorithms (supervised and semisupervised) are employed with decision- and feature-level fusion to model the attack detection problem. The relationships between statistical and geometric properties of attack vectors employed in the attack scenarios and learning algorithms are analyzed to detect unobservable attacks using statistical learning methods. The proposed algorithms are examined on various IEEE test systems. Experimental analyses show that machine learning algorithms can detect attacks with performances higher than attack detection algorithms that employ state vector estimation methods in the proposed attack detection framework.

Building Detection With Decision Fusion

A novel decision fusion approach to building detection problem in VHR optical satellite images is proposed. The method combines the detection results of multiple classifiers under a hierarchical architecture, called Fuzzy Stacked Generalization (FSG). After an initial segmentation and pre-processing step, a large variety of color, texture and shape features are extracted from each segment. Then, the segments, represented in K different feature spaces are classified by K different base-layer classifiers of the FSG architecture. The class membership values of the segments, which represent the decisions of different base-layer classifiers in a decision space, are aggregated to form a fusion space which is then fed to a meta-layer classifier of the FSG to label the vectors in the fusion space. The paper presents the performance results of the proposed decision fusion model by a comparison with the state of the art machine learning algorithms. The results show that fusing the decisions of multiple classifiers improves the performance, when they are ensembled under the suggested hierarchical learning architecture.

A perceptual shape descriptor

We represent a two dimensional object silhouette by a one dimensional descriptor, which preserves the perceptual structure of its shape. The proposed descriptor is based on the moments of the angles between the bearings of a point on the boundary, in a set of neighborhood systems. At each point on the boundary, the angle between a pair of bearings is calculated to extract the topological information of the boundary in a given locality. The proposed method does not use any heuristic rule or empirical threshold value in the shape representation. The similarity between the patterns is measured by elastic matching of the descriptors. The proposed shape descriptor is tested on dataset of MPEG 7 Core Experiments Shape-1. The experiments show better results than the previous studies reported in the literature.

Functional Mesh Learning for pattern analysis of cognitive processes

We propose a statistical learning model for classifying cognitive processes based on distributed patterns of neural activation in the brain, acquired via functional magnetic resonance imaging (fMRI). In the proposed learning machine, local meshes are formed around each voxel. The distance between voxels in the mesh is determined by using functional neighborhood concept. In order to define functional neighborhood, the similarities between the time series recorded for voxels are measured and functional connectivity matrices are constructed. Then, the local mesh for each voxel is formed by including the functionally closest neighboring voxels in the mesh. The relationship between the voxels within a mesh is estimated by using a linear regression model. These relationship vectors, called Functional Connectivity aware Local Relational Features (FC-LRF) are then used to train a statistical learning machine. The proposed method was tested on a recognition memory experiment, including data pertaining to encoding and retrieval of words belonging to ten different semantic categories. Two popular classifiers, namely k-Nearest Neighbor and Support Vector Machine, are trained in order to predict the semantic category of the item being retrieved, based on activation patterns during encoding. The classification performance of the Functional Mesh Learning model, which range in 62-68% is superior to the classical multi-voxel pattern analysis (MVPA) methods, which range in 40-48%, for ten semantic categories.

Representation Learning for Contextual Object and Region Detection in Remote Sensing

The performance of object recognition and classification on remote sensing imagery is highly dependent on the quality of extracted features, amount of labelled data and the priors defined for contextual models. In this study, we examine the representation learning opportunities for remote sensing. First we attacked localization of contextual cues for complex object detection using disentangling factors learnt from a small amount of labelled data. The complex object, which consists of several sub-parts is further represented under the Conditional Markov Random Fields framework. As a second task, end-to-end target detection using convolutional sparse auto-encoders (CSA) using large amount of unlabelled data is analysed. Proposed methodologies are tested on complex airfield detection problem using Conditional Random Fields and recognition of dispersal areas, park areas, taxi routes, airplanes using CSA. The method is also tested on the detection of the dry docks in harbours. Performance of the proposed method is compared with standard feature engineering methods and found competitive with currently used rule-based and supervised methods.

On the Performance of Stacked Generalization Classifiers

Stacked Generalization (SG) is an ensemble learning technique, which aims to increase the performance of individual classifiers by combining them under a hierarchical architecture. In many applications, this technique performs better than the individual classifiers. However, in some applications, the performance of the technique goes astray, for the reasons that are not well-known. In this work, the performance of Stacked Generalization technique is analyzed with respect to the performance of the individual classifiers under the architecture. This work shows that the success of the SG highly depends on how the individual classifiers share to learn the training set, rather than the performance of the individual classifiers. The experiments explore the learning mechanisms of SG to achieve the high performance. The relationship between the performance of the individual classifiers and that of SG is also investigated.

Modeling Voxel Connectivity for Brain Decoding

The massively dynamic nature of human brain cannot be represented by considering only a collection of voxel intensity values obtained from fMRI measurements. It has been observed that the degree of connectivity among voxels provide important information for modeling cognitive activities. Moreover, spatially close voxels act together to generate similar BOLD responses to the same stimuli. In this study, we propose a local mesh model, called Local Mesh Model with Temporal Measurements (LMM-TM), to first estimate spatial relationship among a set of voxels using spatial and temporal data measured at each voxel, and then employ the relationship for the construction of a connectivity model for brain decoding. For this purpose, we first construct a local mesh around each voxel (called seed voxel) by connecting it to its spatially nearest neighbors. Then, we represent the BOLD response of each seed voxel in terms of linear combination of the BOLD responses of its p-nearest neighbors. The relationship between a seed voxel and its neighbors is estimated by solving a linear regression problem. The estimated mesh arc weights are used to model local connectivity among the voxels that reside in a spatial neighborhood. Using these weights as features, we train Support Vector Machines and k-Nearest Neighbor classifiers. We test our model on a visual object recognition experiment. In the experimental analysis, we observe that classifiers that employ our features perform better than classifiers that employ raw voxel intensity values, local mesh model weights and features extracted using distance metrics such as Euclidean distance, cosine similarity and Pearson correlation.

Distributed models for sparse attack construction and state vector estimation in the smart grid

Two distributed attack models and two distributed state vector estimation methods are introduced to handle the sparsity of smart grid networks in order to employ unobservable false data injection attacks and estimate state vectors. First, Distributed Sparse Attacks in which attackers process local measurements in order to achieve consensus for an attack vector are introduced. In the second attack model, called Collective Sparse Attacks, it is assumed that the topological information of the network and the measurements is available to attackers. However, attackers employ attacks to the groups of state vectors. The first distributed state vector estimation method, called Distributed State Vector Estimation, assumes that observed measurements are distributed in groups or clusters in the network. The second method, called Collaborative Sparse State Vector Estimation, consists of different operators estimating subsets of state variables. Therefore, state variables are assumed to be distributed in groups and accessed by the network operators locally. The network operators compute their local estimates and send the estimated values to a centralized network operator in order to update the estimated values.