Complexity of inferences in polytree-shaped semi-qualitative probabilistic networks

Semi-qualitative probabilistic networks (SQPNs) merge two important graphical model formalisms: Bayesian networks and qualitative probabilistic networks. They provade a very Complexity of inferences in polytree-shaped semi-qualitative probabilistic networks and qualitative probabilistic networks. They provide a very general modeling framework by allowing the combination of numeric and qualitative assessments over a discrete domain, and can be compactly encoded by exploiting the same factorization of joint probability distributions that are behind the bayesian networks. This paper explores the computational complexity of semi-qualitative probabilistic networks, and takes the polytree-shaped networks as its main target. We show that the inference problem is coNP-Complete for binary polytrees with multiple observed nodes. We also show that interferences can be performed in time linear in the number of nodes if there is a single observed node. Because our proof is construtive, we obtain an efficient linear time algorithm for SQPNs under such assumptions. To the best of our knowledge, this is the first exact polynominal-time algorithm for SQPn. Together these results provide a clear picture of the inferential complexity in polytree-shaped SQPNs.

Hierarchical multi-label classification using local neural networks

Hierarchical multi-label classification is a complex classification task where the classes involved in the problem are hierarchically structured and each example may simultaneously belong to more than one class in each hierarchical level. In this paper, we extend our previous works, where we investigated a new local-based classification method that incrementally trains a multi-layer perceptron for each level of the classification hierarchy. Predictions made by a neural network in a given level are used as inputs to the neural network responsible for the prediction in the next level. We compare the proposed method with one state-of-the-art decision-tree induction method and two decision-tree induction methods, using several hierarchical multi-label classification datasets. We perform a thorough experimental analysis, showing that our method obtains competitive results to a robust global method regarding both precision and recall evaluation measures.

Netzwerke von Nervenzellen auf strukturierten Oberflächen charakterisiert mit optischen und elektrophysiologischen Methoden

Das Wachstum von Nervenzellen und deren Verbindungen im zentralen und peripheren Nervensystem wird durch Proteine der extrazellulären Matrix kontrolliert. In dieser Arbeit wurde das Matrixprotein Laminin verwendet, um Netzwerke von Nervenzellen auf künstlichen Substraten in vitro zu erzeugen. Zu diesem Zweck wurden Lamininstrukturen mit Mikrostempeln aus Polydimethylsiloxan auf Zellkultursubstrate übertragen. Die Mikrostempel wurden in einem mehrstufigen Verfahren durch Abformung von photolithographisch hergestellten Masken angefertigt. Nach Vorversuchen mit neuronal differenzierten Zellen der Zellinien MzN und P19 zur Identifizierung geeigneter Abmessungen der Mikrotrukturen, gelang die Realisierung von Linien- und Gitternetzwerken sowie von komplexeren Schaltungen. Eine morphologische Charakterisierung der erzeugten Netzwerke erfolgte durch Phasenkontrast- und Fluoreszenzmikroskopie.Elektrophysiologische Messungen wurden mit der Patch-Clamp Technik an einer Kultur von Nervenzellen aus primär isolierten Hirnschnitten durchgeführt. Der Erhalt des intakten Zellverbundes im Hirnschnitt sollte Bedingungen möglichst nahe zur Situation in vivo schaffen, um die Bildung von Synapsen zu begünstigen. In Patch-Clamp Messungen an bis zu drei Neuronen gleichzeitig, gelang der Nachweis synaptischer Kopplung in strukturierten Netzwerken solcher Hirnschnitt-Kulturen. Sowohl funktionale chemische Synapsen, als auch Ohm'sche Kopplung über Gap-Junctions wurde beobachtet. Es wurde ein elektrisches Kopplungsmodell abgeleitet. Die Signalleitung in den Nervenfasern erfolgt demnach wie in einem zylindrischen, durch die Zellmembran von der Umgebung isolierten Kabel.

Nondestructive evaluation of concrete compression strength by means of Artificial Neural Network (ANN)

The evaluation of structural performance of existing concrete buildings, built according to standards and materials quite different to those available today, requires procedures and methods able to cover lack of data about mechanical material properties and reinforcement detailing. To this end detailed inspections and test on materials are required. As a consequence tests on drilled cores are required; on the other end, it is stated that non-destructive testing (NDT) cannot be used as the only mean to get structural information, but can be used in conjunction with destructive testing (DT) by a representative correlation between DT and NDT. The aim of this study is to verify the accuracy of some formulas of correlation available in literature between measured parameters, i.e. rebound index, ultrasonic pulse velocity and compressive strength (SonReb Method). To this end a relevant number of DT and NDT tests has been performed on many school buildings located in Cesena (Italy). The above relationships have been assessed on site correlating NDT results to strength of core drilled in adjacent locations. Nevertheless, concrete compressive strength assessed by means of NDT methods and evaluated with correlation formulas has the advantage of being able to be implemented and used for future applications in a much more simple way than other methods, even if its accuracy is strictly limited to the analysis of concretes having the same characteristics as those used for their calibration. This limitation warranted a search for a different evaluation method for the non-destructive parameters obtained on site. To this aim, the methodology of neural identification of compressive strength is presented. Artificial Neural Network (ANN) suitable for the specific analysis were chosen taking into account the development presented in the literature in this field. The networks were trained and tested in order to detect a more reliable strength identification methodology.

Deep learning for computer vision: a comparison between convolutional neural networks and hierarchical temporal memories on object recognition tasks

In recent years, Deep Learning techniques have shown to perform well on a large variety of problems both in Computer Vision and Natural Language Processing, reaching and often surpassing the state of the art on many tasks. The rise of deep learning is also revolutionizing the entire field of Machine Learning and Pattern Recognition pushing forward the concepts of automatic feature extraction and unsupervised learning in general. However, despite the strong success both in science and business, deep learning has its own limitations. It is often questioned if such techniques are only some kind of brute-force statistical approaches and if they can only work in the context of High Performance Computing with tons of data. Another important question is whether they are really biologically inspired, as claimed in certain cases, and if they can scale well in terms of "intelligence". The dissertation is focused on trying to answer these key questions in the context of Computer Vision and, in particular, Object Recognition, a task that has been heavily revolutionized by recent advances in the field. Practically speaking, these answers are based on an exhaustive comparison between two, very different, deep learning techniques on the aforementioned task: Convolutional Neural Network (CNN) and Hierarchical Temporal memory (HTM). They stand for two different approaches and points of view within the big hat of deep learning and are the best choices to understand and point out strengths and weaknesses of each of them. CNN is considered one of the most classic and powerful supervised methods used today in machine learning and pattern recognition, especially in object recognition. CNNs are well received and accepted by the scientific community and are already deployed in large corporation like Google and Facebook for solving face recognition and image auto-tagging problems. HTM, on the other hand, is known as a new emerging paradigm and a new meanly-unsupervised method, that is more biologically inspired. It tries to gain more insights from the computational neuroscience community in order to incorporate concepts like time, context and attention during the learning process which are typical of the human brain. In the end, the thesis is supposed to prove that in certain cases, with a lower quantity of data, HTM can outperform CNN.

THE RESONATE-AND-FIRE NEURON: TIME DEPENDENT AND FREQUENCY SELECTIVE NEURONS IN NEURAL NETWORKS

The means through which the nervous system perceives its environment is one of the most fascinating questions in contemporary science. Our endeavors to comprehend the principles of neural science provide an instance of how biological processes may inspire novel methods in mathematical modeling and engineering. The application ofmathematical models towards understanding neural signals and systems represents a vibrant field of research that has spanned over half a century. During this period, multiple approaches to neuronal modeling have been adopted, and each approach is adept at elucidating a specific aspect of nervous system function. Thus while bio-physical models have strived to comprehend the dynamics of actual physical processes occurring within a nerve cell, the phenomenological approach has conceived models that relate the ionic properties of nerve cells to transitions in neural activity. Further-more, the field of neural networks has endeavored to explore how distributed parallel processing systems may become capable of storing memory. Through this project, we strive to explore how some of the insights gained from biophysical neuronal modeling may be incorporated within the field of neural net-works. We specifically study the capabilities of a simple neural model, the Resonate-and-Fire (RAF) neuron, whose derivation is inspired by biophysical neural modeling. While reflecting further biological plausibility, the RAF neuron is also analytically tractable, and thus may be implemented within neural networks. In the following thesis, we provide a brief overview of the different approaches that have been adopted towards comprehending the properties of nerve cells, along with the framework under which our specific neuron model relates to the field of neuronal modeling. Subsequently, we explore some of the time-dependent neurocomputational capabilities of the RAF neuron, and we utilize the model to classify logic gates, and solve the classic XOR problem. Finally we explore how the resonate-and-fire neuron may be implemented within neural networks, and how such a network could be adapted through the temporal backpropagation algorithm.

Reduced neuronal efficacy in progressive mild cognitive impairment: a prospective fMRI study on visuospatial processing

Mild cognitive impairment (MCI) often refers to the preclinical stage of dementia, where the majority develop Alzheimer's disease (AD). Given that neurodegenerative burden and compensatory mechanisms might exist before accepted clinical symptoms of AD are noticeable, the current prospective study aimed to investigate the functioning of brain regions in the visuospatial networks responsible for preclinical symptoms in AD using event-related functional magnetic resonance imaging (fMRI). Eighteen MCI patients were evaluated and clinically followed for approximately 3 years. Five progressed to AD (PMCI) and eight remained stable (SMCI). Thirteen age-, gender- and education-matched controls also participated. An angle discrimination task with varying task demands was used. Brain activation patterns as well as task demand-dependent and -independent signal changes between the groups were investigated by using an extended general linear model including individual performance (reaction time [RT]) of each single trial. Similar behavioral (RT and accuracy) responses were observed between MCI patients and controls. A network of bilateral activations, e.g. dorsal pathway, which increased linearly with increasing task demand, was engaged in all subjects. Compared with SMCI patients and controls, PMCI patients showed a stronger relation between task demand and brain activity in left superior parietal lobules (SPL) as well as a general task demand-independent increased activation in left precuneus. Altered brain function can be detected at a group level in individuals that progress to AD before changes occur at the behavioral level. Increased parietal activation in PMCI could reflect a reduced neuronal efficacy due to accumulating AD pathology and might predict future clinical decline in patients with MCI.

Neural Networks as Cybernetic Systems - Part I

Neural Networks as Cybernetic Systems is a textbox that combines classical systems theory with artificial neural network technology.

Neural Networks as Cybernetic Systems - Part II

Neural Networks as Cybernetic Systems is a textbox that combines classical systems theory with artificial neural network technology.

Neural Networks as Cybernetic Systems (3rd and revised edition) - Part I

Neural Networks as Cybernetic Systems is a textbox that combines classical systems theory with artificial neural network technology. This third edition essentially compares with the 2nd one, but has been improved by correction of errors and by a rearrangement and minor expansion of the sections referring to recurrent networks. These changes hopefully allow for an easier comprehension of the essential aspects of this important domain that has received growing attention during the last years.

Neural Networks as Cybernetic Systems (3rd and revised edition) - Part 2

eural Networks as Cybernetic Systems is a textbox that combines classical systems theory with artificial neural network technology. This third edition essentially compares with the 2nd one, but has been improved by correction of errors and by a rearrangement and minor expansion of the sections referring to recurrent networks. These changes hopefully allow for an easier comprehension of the essential aspects of this important domain that has received growing attention during the last years.

EEG microstates associated with salience and frontoparietal networks in frontotemporal dementia, schizophrenia and Alzheimer's disease

OBJECTIVE: There are relevant links between resting-state fMRI networks, EEG microstate classes and psychopathological alterations in mental disorders associated with frontal lobe dysfunction. We hypothesized that a certain microstate class, labeled C and correlated with the salience network, was impaired early in frontotemporal dementia (FTD), and that microstate class D, correlated with the frontoparietal network, was impaired in schizophrenia. METHODS: We measured resting EEG microstate parameters in patients with mild FTD (n = 18), schizophrenia (n = 20), mild Alzheimer's disease (AD; n = 19) and age-matched controls (old n = 19, young n = 18) to investigate neuronal dynamics at the whole-brain level. RESULTS: The duration of class C was significantly shorter in FTD than in controls and AD, and the duration of class D was significantly shorter in schizophrenia than in controls, FTD and AD. Transition analysis showed a reversed sequence of activation of classes C and D in FTD and schizophrenia patients compared with that in controls, with controls preferring transitions from C to D, and patients preferring D to C. CONCLUSION: The duration and sequence of EEG microstates reflect specific aberrations of frontal lobe functions in FTD and schizophrenia. SIGNIFICANCE: This study highlights the importance of subsecond brain dynamics for understanding of psychiatric disorders.

Investigation on the evolution of an indoor robotic localization system based on wireless networks

This work addresses the evolution of an artificial neural network (ANN) to assist in the problem of indoor robotic localization. We investigate the design and building of an autonomous localization system based on information gathered from wireless networks (WN). The article focuses on the evolved ANN, which provides the position of a robot in a space, as in a Cartesian coordinate system, corroborating with the evolutionary robotic research area and showing its practical viability. The proposed system was tested in several experiments, evaluating not only the impact of different evolutionary computation parameters but also the role of the transfer functions on the evolution of the ANN. Results show that slight variations in the parameters lead to significant differences on the evolution process and, therefore, in the accuracy of the robot position.