On-line learning in neural networks

On-line learning is one of the most powerful and commonly used techniques for training large layered networks and has been used successfully in many real-world applications. Traditional analytical methods have been recently complemented by ones from statistical physics and Bayesian statistics. This powerful combination of analytical methods provides more insight and deeper understanding of existing algorithms and leads to novel and principled proposals for their improvement. This book presents a coherent picture of the state-of-the-art in the theoretical analysis of on-line learning. An introduction relates the subject to other developments in neural networks and explains the overall picture. Surveys by leading experts in the field combine new and established material and enable non-experts to learn more about the techniques and methods used. This book, the first in the area, provides a comprehensive view of the subject and will be welcomed by mathematicians, scientists and engineers, whether in industry or academia.

Using non-parametric search algorithms to forecast daily excess stock returns

Are the learning procedures of genetic algorithms (GAs) able to generate optimal architectures for artificial neural networks (ANNs) in high frequency data? In this experimental study,GAs are used to identify the best architecture for ANNs. Additional learning is undertaken by the ANNs to forecast daily excess stock returns. No ANN architectures were able to outperform a random walk,despite the finding of non-linearity in the excess returns. This failure is attributed to the absence of suitable ANN structures and further implies that researchers need to be cautious when making inferences from ANN results that use high frequency data.

Visual data mining: integrating machine learning with information visualization

Today, the data available to tackle many scientific challenges is vast in quantity and diverse in nature. The exploration of heterogeneous information spaces requires suitable mining algorithms as well as effective visual interfaces. Most existing systems concentrate either on mining algorithms or on visualization techniques. Though visual methods developed in information visualization have been helpful, for improved understanding of a complex large high-dimensional dataset, there is a need for an effective projection of such a dataset onto a lower-dimension (2D or 3D) manifold. This paper introduces a flexible visual data mining framework which combines advanced projection algorithms developed in the machine learning domain and visual techniques developed in the information visualization domain. The framework follows Shneiderman’s mantra to provide an effective user interface. The advantage of such an interface is that the user is directly involved in the data mining process. We integrate principled projection methods, such as Generative Topographic Mapping (GTM) and Hierarchical GTM (HGTM), with powerful visual techniques, such as magnification factors, directional curvatures, parallel coordinates, billboarding, and user interaction facilities, to provide an integrated visual data mining framework. Results on a real life high-dimensional dataset from the chemoinformatics domain are also reported and discussed. Projection results of GTM are analytically compared with the projection results from other traditional projection methods, and it is also shown that the HGTM algorithm provides additional value for large datasets. The computational complexity of these algorithms is discussed to demonstrate their suitability for the visual data mining framework.

An investigation into inductive parameter learning in complex hierarchical knowledge structures representing clinical expertise

This dissertation investigates the very important and current problem of modelling human expertise. This is an apparent issue in any computer system emulating human decision making. It is prominent in Clinical Decision Support Systems (CDSS) due to the complexity of the induction process and the vast number of parameters in most cases. Other issues such as human error and missing or incomplete data present further challenges. In this thesis, the Galatean Risk Screening Tool (GRiST) is used as an example of modelling clinical expertise and parameter elicitation. The tool is a mental health clinical record management system with a top layer of decision support capabilities. It is currently being deployed by several NHS mental health trusts across the UK. The aim of the research is to investigate the problem of parameter elicitation by inducing them from real clinical data rather than from the human experts who provided the decision model. The induced parameters provide an insight into both the data relationships and how experts make decisions themselves. The outcomes help further understand human decision making and, in particular, help GRiST provide more accurate emulations of risk judgements. Although the algorithms and methods presented in this dissertation are applied to GRiST, they can be adopted for other human knowledge engineering domains.

Exploring ant-based algorithms for gene expression data analysis

Objective: Recently, much research has been proposed using nature inspired algorithms to perform complex machine learning tasks. Ant colony optimization (ACO) is one such algorithm based on swarm intelligence and is derived from a model inspired by the collective foraging behavior of ants. Taking advantage of the ACO in traits such as self-organization and robustness, this paper investigates ant-based algorithms for gene expression data clustering and associative classification. Methods and material: An ant-based clustering (Ant-C) and an ant-based association rule mining (Ant-ARM) algorithms are proposed for gene expression data analysis. The proposed algorithms make use of the natural behavior of ants such as cooperation and adaptation to allow for a flexible robust search for a good candidate solution. Results: Ant-C has been tested on the three datasets selected from the Stanford Genomic Resource Database and achieved relatively high accuracy compared to other classical clustering methods. Ant-ARM has been tested on the acute lymphoblastic leukemia (ALL)/acute myeloid leukemia (AML) dataset and generated about 30 classification rules with high accuracy. Conclusions: Ant-C can generate optimal number of clusters without incorporating any other algorithms such as K-means or agglomerative hierarchical clustering. For associative classification, while a few of the well-known algorithms such as Apriori, FP-growth and Magnum Opus are unable to mine any association rules from the ALL/AML dataset within a reasonable period of time, Ant-ARM is able to extract associative classification rules.

Visual data mining using principled projection algorithms and information visualization techniques

We introduce a flexible visual data mining framework which combines advanced projection algorithms from the machine learning domain and visual techniques developed in the information visualization domain. The advantage of such an interface is that the user is directly involved in the data mining process. We integrate principled projection algorithms, such as generative topographic mapping (GTM) and hierarchical GTM (HGTM), with powerful visual techniques, such as magnification factors, directional curvatures, parallel coordinates and billboarding, to provide a visual data mining framework. Results on a real-life chemoinformatics dataset using GTM are promising and have been analytically compared with the results from the traditional projection methods. It is also shown that the HGTM algorithm provides additional value for large datasets. The computational complexity of these algorithms is discussed to demonstrate their suitability for the visual data mining framework. Copyright 2006 ACM.

Automating Exersises Validation in e-Learning Environments

E-learning is supposing an innovation in teaching, raising from the development of new technologies. It is based in a set of educational resources, including, among others, multimedia or interactive contents accessible through Internet or Intranet networks. A whole spectrum of tools and services support e-learning, some of them include auto-evaluation and automated correction of test-like exercises, however, this sort of exercises are very constrained because of its nature: fixed contents and correct answers suppose a limit in the way teachers may evaluation students. In this paper we propose a new engine that allows validating complex exercises in the area of Data Structures and Algorithms. Correct solutions to exercises do not rely only in how good the execution of the code is, or if the results are same as expected. A set of criteria on algorithm complexity or correctness in the use of the data structures are required. The engine presented in this work covers a wide set of exercises with these characteristics allowing teachers to establish the set of requirements for a solution, and students to obtain a measure on the quality of their solution in the same terms that are later required for exams.

Complex Adaptive On-line Learning System “Theory of Automatic Control"

The controlled from distance teaching (DT) in the system of technical education has a row of features: complication of informative content, necessity of development of simulation models and trainers for conducting of practical and laboratory employments, conducting of knowledge diagnostics on the basis of mathematical-based algorithms, organization of execution collective projects of the applied setting. For development of the process of teaching bases of fundamental discipline control system Theory of automatic control (TAC) the combined approach of optimum combination of existent programmatic instruments of support was chosen DT and own developments. The system DT TAC included: controlled from distance course (DC) of TAC, site of virtual laboratory practical works in LAB.TAC and students knowledge remote diagnostic system d-tester.

An eMathTeacher Tool for Active Learning Fleury’s Algorithm

An eMathTeacher [Sánchez-Torrubia 2007a] is an eLearning on line self assessment tool that help students to active learning math algorithms by themselves, correcting their mistakes and providing them with clues to find the right solution. The tool presented in this paper is an example of this new concept on Computer Aided Instruction (CAI) resources and has been implemented as a Java applet and designed as an auxiliary instrument for both classroom teaching and individual practicing of Fleury’s algorithm. This tool, included within a set of eMathTeacher tools, has been designed as educational complement of Graph Algorithm active learning for first course students. Its characteristics of visualization, simplicity and interactivity, make this tutorial a great value pedagogical instrument.

Online learning in discrete hidden Markov models

We present and analyze three different online algorithms for learning in discrete Hidden Markov Models (HMMs) and compare their performance with the Baldi-Chauvin Algorithm. Using the Kullback-Leibler divergence as a measure of the generalization error we draw learning curves in simplified situations and compare the results. The performance for learning drifting concepts of one of the presented algorithms is analyzed and compared with the Baldi-Chauvin algorithm in the same situations. A brief discussion about learning and symmetry breaking based on our results is also presented. © 2006 American Institute of Physics.

A Planning Model with Resources in e-Learning

This work proposes a model for planning of education based on resources and layers. Each learning material or concept is determined by certain characteristics: a layer and a list of resources and resource values. Models of studied subject domain, learner, information and verification unit, learning material, plan of education and education have been defined. The plan of education can be conventional, statical, author’s and dynamic. Algorithms for course generation, dynamic plan generation and carrying out education are presented. The proposed model for planning of education based on resources and layers has been included in the system PeU.

Cluster Analysis of Some Clases of Objects by Application of the Test Recognition Algorithms

The paper treats the task for cluster analysis of a given assembly of objects on the basis of the information contained in the description table of these objects. Various methods of cluster analysis are briefly considered. Heuristic method and rules for classification of the given assembly of objects are presented for the cases when their division into classes and the number of classes is not known. The algorithm is checked by a test example and two program products (PP) – learning systems and software for company management. Analysis of the results is presented.

Algorithms for super-resolution of images based on Sparse Representation and Manifolds

lmage super-resolution is defined as a class of techniques that enhance the spatial resolution of images. Super-resolution methods can be subdivided in single and multi image methods. This thesis focuses on developing algorithms based on mathematical theories for single image super­ resolution problems. lndeed, in arder to estimate an output image, we adopta mixed approach: i.e., we use both a dictionary of patches with sparsity constraints (typical of learning-based methods) and regularization terms (typical of reconstruction-based methods). Although the existing methods already per- form well, they do not take into account the geometry of the data to: regularize the solution, cluster data samples (samples are often clustered using algorithms with the Euclidean distance as a dissimilarity metric), learn dictionaries (they are often learned using PCA or K-SVD). Thus, state-of-the-art methods still suffer from shortcomings. In this work, we proposed three new methods to overcome these deficiencies. First, we developed SE-ASDS (a structure tensor based regularization term) in arder to improve the sharpness of edges. SE-ASDS achieves much better results than many state-of-the- art algorithms. Then, we proposed AGNN and GOC algorithms for determining a local subset of training samples from which a good local model can be computed for recon- structing a given input test sample, where we take into account the underlying geometry of the data. AGNN and GOC methods outperform spectral clustering, soft clustering, and geodesic distance based subset selection in most settings. Next, we proposed aSOB strategy which takes into account the geometry of the data and the dictionary size. The aSOB strategy outperforms both PCA and PGA methods. Finally, we combine all our methods in a unique algorithm, named G2SR. Our proposed G2SR algorithm shows better visual and quantitative results when compared to the results of state-of-the-art methods.

Learning and Exploiting Camera Geometry for Computer Vision

This work explores the use of statistical methods in describing and estimating camera poses, as well as the information feedback loop between camera pose and object detection. Surging development in robotics and computer vision has pushed the need for algorithms that infer, understand, and utilize information about the position and orientation of the sensor platforms when observing and/or interacting with their environment.

The first contribution of this thesis is the development of a set of statistical tools for representing and estimating the uncertainty in object poses. A distribution for representing the joint uncertainty over multiple object positions and orientations is described, called the mirrored normal-Bingham distribution. This distribution generalizes both the normal distribution in Euclidean space, and the Bingham distribution on the unit hypersphere. It is shown to inherit many of the convenient properties of these special cases: it is the maximum-entropy distribution with fixed second moment, and there is a generalized Laplace approximation whose result is the mirrored normal-Bingham distribution. This distribution and approximation method are demonstrated by deriving the analytical approximation to the wrapped-normal distribution. Further, it is shown how these tools can be used to represent the uncertainty in the result of a bundle adjustment problem.

Another application of these methods is illustrated as part of a novel camera pose estimation algorithm based on object detections. The autocalibration task is formulated as a bundle adjustment problem using prior distributions over the 3D points to enforce the objects' structure and their relationship with the scene geometry. This framework is very flexible and enables the use of off-the-shelf computational tools to solve specialized autocalibration problems. Its performance is evaluated using a pedestrian detector to provide head and foot location observations, and it proves much faster and potentially more accurate than existing methods.

Finally, the information feedback loop between object detection and camera pose estimation is closed by utilizing camera pose information to improve object detection in scenarios with significant perspective warping. Methods are presented that allow the inverse perspective mapping traditionally applied to images to be applied instead to features computed from those images. For the special case of HOG-like features, which are used by many modern object detection systems, these methods are shown to provide substantial performance benefits over unadapted detectors while achieving real-time frame rates, orders of magnitude faster than comparable image warping methods.

The statistical tools and algorithms presented here are especially promising for mobile cameras, providing the ability to autocalibrate and adapt to the camera pose in real time. In addition, these methods have wide-ranging potential applications in diverse areas of computer vision, robotics, and imaging.

Structure learning of context-specific graphical models

Abstract The ultimate problem considered in this thesis is modeling a high-dimensional joint distribution over a set of discrete variables. For this purpose, we consider classes of context-specific graphical models and the main emphasis is on learning the structure of such models from data. Traditional graphical models compactly represent a joint distribution through a factorization justi ed by statements of conditional independence which are encoded by a graph structure. Context-speci c independence is a natural generalization of conditional independence that only holds in a certain context, speci ed by the conditioning variables. We introduce context-speci c generalizations of both Bayesian networks and Markov networks by including statements of context-specific independence which can be encoded as a part of the model structures. For the purpose of learning context-speci c model structures from data, we derive score functions, based on results from Bayesian statistics, by which the plausibility of a structure is assessed. To identify high-scoring structures, we construct stochastic and deterministic search algorithms designed to exploit the structural decomposition of our score functions. Numerical experiments on synthetic and real-world data show that the increased exibility of context-specific structures can more accurately emulate the dependence structure among the variables and thereby improve the predictive accuracy of the models.