Functional MRI (fMRI) is a valuable brain imaging technique. A significant problem, when analyzing fMRI time series, is the finding of functional brain networks when the brain is at rest, i.e. no external stimulus is applied to the subject. In this work, we present a probabilistic method to estimate the Resting State Networks (RSNs) using a model-based approach. More specifically, RSNs are assumed to be the result of a clustering procedure. In order to perform the clustering, a mixture of regression models are used. Furthermore, special care has been given in order to incorporate important functionalities, such as spatial and embedded sparsity enforcing properties, through the use of informative priors over the model parameters. Another interesting feature of the proposed scheme is the flexibility to handle all the brain time series at once, allowing more robust solutions. We provide comparative experimental results, using an artificial fMRI dataset and two real resting state fMRI datasets, that empirically illustrate the efficiency of the proposed regression mixture model.

In this paper, we introduce AngryBER, an intelligent agent architecture on the Angry Birds domain that employs a Bayesian ensemble inference mechanism to promote decision-making abilities. It is based on an efficient tree-like structure for encoding and representing game screenshots, where it exploits its enhanced modeling capabilities. This has the advantage to establish an informative feature space and translate the task of game playing into a regression analysis problem. A Bayesian ensemble regression framework is presented by considering that every combination of objects' material and bird type has its own regression model. We address the problem of action selection as a multiarmed bandit problem, where the upper confidence bound (UCB) strategy has been used. An efficient online learning procedure has been also developed for training the regression models. We have evaluated the proposed methodology on several game levels, and compared its performance with published results of all agents that participated in the 2013 and 2014 Angry Birds AI competitions. The superiority of the new method is readily deduced by inspecting the reported results.

Functional MRI (fMRI) is one of the most important techniques to study the human brain. A relatively new problem to the analysis of fMRI data is the identification of brain networks when the brain is at rest i.e. no external stimulus is applied to the subject. In this work a method to find the Resting State Networks (RSNs), using fMRI time series, is proposed. To achieve that our method uses the Regression Mixtures Models (RMMs). RMMs are mixture models specifically design to cluster time series. Furthermore, our method takes into account the spatial correlations of fMRI data by using a new functional for the responsibilities of the mixture. Experimental results have showed the usefullness of the proposed approach compared to other methods of the field such as the k-means algorithm.

In this article, we introduce Newtonian spectral clustering, a method that employs Newtonian preprocessing to promote cluster perspicuity and trajectory analysis to gain valuable affinity information. A simple two-body potential is used to model the interaction under the influence of which the points move according to Newton's second law. This procedure produces a transformed data set with reduced cluster overlap, which favors the spectral clustering approach. This is so, because the affinity matrix can be enriched with information derived from the underlying interaction model. Special care is also given to estimate the Gaussian kernel parameter, since its role is important for the clustering procedure. The method is further extended appropriately to treat problems of high dimensionality. We have tested the proposed methodology on several benchmark data and compared its performance to that of rival techniques. The superiority of the new approach is readily deduced by inspecting the reported results.

Functional magnetic resonance imaging (fMRI) has become one of the most important techniques for studying the human brain in action. A common problem in fMRI analysis is the detection of activated brain regions in response to an experimental task. In this work we propose a novel clustering approach for addressing this issue using an adaptive regression mixture model. The main contribution of our method is the employment of both spatial and sparse properties over the body of the mixture model. Thus, the clustering approach is converted into a maximum a posteriori estimation approach, where the expectation-maximization algorithm is applied for model training. Special care is also given to estimate the kernel scalar parameter per cluster of the design matrix by presenting a multi-kernel scheme. In addition an incremental training procedure is presented so as to make the approach independent on the initialization of the model parameters. The latter also allows us to introduce an efficient stopping criterion of the process for determining the optimum brain activation area. To assess the effectiveness of our method, we have conducted experiments with simulated and real fMRI data, where we have demonstrated its ability to produce improved performance and functional activation detection capabilities.

We present a new regression mixture model where each mixture component is a multi-kernel version of the Relevance Vector Machine (RVM). In the proposed model, we exploit the enhanced modeling capability of RVMs due to their embedded sparsity enforcing properties. %The main contribution of this %work is the employment of RVM models as components of a mixture %model and their application to the time series clustering problem. Moreover, robustness is achieved with respect to the kernel parameters, by employing a weighted multi-kernel scheme. The mixture model is trained using the maximum a posteriori (MAP) approach, where the Expectation Maximization (EM) algorithm is applied offering closed form update equations for the model parameters. An incremental learning methodology is also presented to tackle the parameter initialization problem of the EM algorithm. The efficiency of the proposed mixture model is empirically demonstrated on the time series clustering problem using various artificial and real benchmark datasets and by performing comparisons with other regression mixture models.

A significant issue in representing reinforcement learning agents in Markov decision processes is how to design efficient feature spaces in order to estimate optimal policy. This particular study addresses this challenge by proposing a compact framework that employs an on-line clustering approach for constructing appropriate basis functions. Also, it performs a state-action trajectory analysis to gain valuable affinity information among clusters and estimate their transition dynamics. Value function approximation is used for policy evaluation in a least-squares temporal difference framework. The proposed method is evaluated in several simulated and real environments, where we took promising results.

In this work we present an advanced Bayesian formulation to the task of control learning that employs the Relevance Vector Machines (RVM) generative model for value function evaluation. The key aspect of the proposed method is the design of the discount return as a generalized linear model that constitutes a well-known probabilistic approach. This allows to augment the model with advantageous sparse priors provided by the RVM's regression framework. We have also taken into account the significant issue of selecting the proper parameters of the kernel design matrix. Experiments have shown that our method produces improved performance in both simulated and real test environments.

In this paper we present a new approach for curve clustering designed for analysis of spatiotemporal data. Such kind of data contains both spatial and temporal patterns that we desire to capture. The proposed methodology is based on regression and Gaussian mixture modeling and the novelty of the herein work is the incorporation of spatial smoothness constraints in the form of a prior for the data labels. This enables the proposed model to take into account the underlying property of spatiotemporal data that spatially adjacent data points most likely should belong to the same cluster. A maximum a posteriori Expectation Maximization (MAP-EM) algorithm is used for learning this model. We present numerical experiments with simulated data where the ground truth is known in order to assess the value of the introduced smoothness constraint, and also with real cardiac perfusion MRI data. The results are very promising and demonstrate the value of the proposed constraint for analysis of such data .

In this paper we present new detectors for additive watermarks when the power of the watermark is unknown. These detectors are based on modeling the image using student-t statistics. As a result, due to the generative properties of the student-t density function, such models are spatially adaptive and the Expectation-Maximization algorithm can be used to obtain maximum likelihood estimates of their parameters. Using these image models detectors based on the generalized likelihood ratio and Rao tests are derived for this problem. Numerical experiments are presented that demonstrate the properties of these detectors and compared them with previously proposed detectors.

Pattern-based Java bytecode compression techniques rely on the identification of identical instruction sequences that occur more than once. Each occurrence of such a sequence is substituted by a single instruction. The sequence defines a pattern that is used for extending the standard bytecode instruction set with the instruction that substitutes the pattern occurrences in the original bytecode. Alternatively, the pattern may be stored in a dictionary that serves for the bytecode decompression. In this case, the instruction that substitutes the pattern in the original bytecode serves as an index to the dictionary. In this paper, we investigate a bytecode compression technique that considers a more general case of patterns. Specifically, we employ the use of an advanced pattern discovery technique that allows locating patterns of an arbitrary length, which may contain a variable number of wildcards in place of certain instruction opcodes or operands. We evaluate the benefits and the limitations of this technique in various scenarios that aim at compressing the reference implementation of MIDP, a standard Java environment for the development of applications for mobile devices.

In this paper, we propose a new methodology for analysis of microarray images. First, a new gridding algorithm is proposed for determining the individual spots and their borders. Then, a Gaussian mixture model (GMM) approach is presented for the analysis of the individual spot images. The main advantages of the proposed methodology are modeling flexibility and adaptability to the data, which are well-known strengths of GMM. The maximum likelihood and maximum a posteriori approaches are used to estimate the GMM parameters via the expectation maximization algorithm. The proposed approach has the ability to detect and compensate for artifacts that might occur in microarray images. This is accomplished by a model-based criterion that selects the number of the mixture components. We present numerical experiments with artificial and real data where we compare the proposed approach with previous ones and existing software tools for microarray image analysis and demonstrate its advantages.

Gaussian mixture models (GMMs) constitute a well-known type of probabilistic neural networks. One of their many successful applications is in image segmentation, where spatially constrained mixture models have been trained using the expectation-maximization (EM) framework. In this letter, we elaborate on this method and propose a new methodology for the M-step of the EM algorithm that is based on a novel constrained optimization formulation. Numerical experiments using simulated images illustrate the superior performance of our method in terms of the attained maximum value of the objective function and segmentation accuracy compared to previous implementations of this approach.

Image processing for analysis of microarray images is an important and challenging problem because imperfections and fabrication artifacts often impair our ability to measure accurately the quantities of interest in these images. In this paper we propose a microarray image analysis framework that provides a new method that automatically addresses each spot area in the image. Then, a new unsupervised clustering method is used which is based on a Gaussian mixture model (GMM) and the minimum description length (MDL) criterion, that allows the automatic spot area segmentation and the image artifacts isolation and correction to obtain more accurate spot quantitative values. Experimental results demonstrates the advantages of the proposed scheme in efficiently analysing microarrays.

MOTIVATION: This paper studies the problem of discovering subsequences, known as motifs, that are common to a given collection of related biosequences, by proposing a greedy algorithm for learning a mixture of motifs model through likelihood maximization. The approach adds sequentially a new motif to a mixture model by performing a combined scheme of global and local search for appropriately initializing its parameters. In addition, a hierarchical partitioning scheme based on kd-trees is presented for partitioning the input dataset in order to speed-up the global searching procedure. The proposed method compares favorably over the well-known MEME approach and treats successfully several drawbacks of MEME. RESULTS: Experimental results indicate that the algorithm is advantageous in identifying larger groups of motifs characteristic of biological families with significant conservation. In addition, it offers better diagnostic capabilities by building more powerful statistical motif-models with improved classification accuracy.

A neural network classifier is presented, which is based on geometrical fuzzy sets. Starting from the construction of the Voronoi diagram of the training patterns, an aggregation of Voronoi regions is performed leading to the identification of larger regions belonging exclusively to one of the pattern classes. The resulting scheme is a constructive algorithm that defines fuzzy clusters of patterns. Based on observations concerning the grade of membership of the training patterns to the created regions, decision probabilities are computed through which the final classification is performed. Experimental results concerning several classification problems indicate that the proposed method achieves high classification rates and compares favorably with other well-known approaches.