Emilio Corchado

A survey of multiple classifier systems as hybrid systems

A current focus of intense research in pattern classification is the combination of several classifier systems, which can be built following either the same or different models and/or datasets building approaches. These systems perform information fusion of classification decisions at different levels overcoming limitations of traditional approaches based on single classifiers. This paper presents an up-to-date survey on multiple classifier system (MCS) from the point of view of Hybrid Intelligent Systems. The article discusses major issues, such as diversity and decision fusion methods, providing a vision of the spectrum of applications that are currently being developed.

Hybrid Artificial Intelligent Systems

This paper deals with discovering frequent sets for quantitative association rules mining with preserved privacy. It focuses on privacy preserving on an individual level, when true individual values, e.g., values of attributes describing customers, are not revealed. Only distorted values and parameters of the distortion procedure are public. However, a miner can discover hidden knowledge, e.g., association rules, from the distorted data. In order to find frequent sets for quantitative association rules mining with preserved privacy, not only does a miner need to discretise continuous attributes, transform them into binary attributes, but also, after both discretisation and binarisation, the calculation of the distortion parameters for new attributes is necessary. Then a miner can apply either MASK (Mining Associations with Secrecy Konstraints) or MMASK (Modified MASK) to find candidates for frequent sets and estimate their supports. In this paper the methodology for calculating distortion parameters of newly created attributes after both discretisation and binarisation of attributes for quantitative association rules mining has been proposed. The new application of MMASK for finding frequent sets in discovering quantitative association rules with preserved privacy has been also presented. The application of MMASK scheme for frequent sets mining in quantitative association rules discovery on real data sets has been experimentally verified. The results of the experiments show that both MASK and MMASK can be applied in frequent sets mining for quantitative association rules with preserved privacy, however, MMASK gives better results in this task.

Intelligent Data Engineering and Automated Learning – IDEAL 2006

Learning and Information Processing -- Data Mining, Retrieval and Management -- Bioinformatics and Bio-inspired Models -- Agents and Hybrid Systems -- Financial Engineering -- Special Session on Nature-Inspired Date Technologies.

Neural visualization of network traffic data for intrusion detection

This study introduces and describes a novel intrusion detection system (IDS) called MOVCIDS (mobile visualization connectionist IDS). This system applies neural projection architectures to detect anomalous situations taking place in a computer network. By its advanced visualization facilities, the proposed IDS allows providing an overview of the network traffic as well as identifying anomalous situations tackled by computer networks, responding to the challenges presented by volume, dynamics and diversity of the traffic, including novel (0-day) attacks. MOVCIDS provides a novel point of view in the field of IDSs by enabling the most interesting projections (based on the fourth order statistics; the kurtosis index) of a massive traffic dataset to be extracted. These projections are then depicted through a functional and mobile visualization interface, providing visual information of the internal structure of the traffic data. The interface makes MOVCIDS accessible from any mobile device to give more accessibility to network administrators, enabling continuous visualization, monitoring and supervision of computer networks. Additionally, a novel testing technique has been developed to evaluate MOVCIDS and other IDSs employing numerical datasets. To show the performance and validate the proposed IDS, it has been tested in different real domains containing several attacks and anomalous situations. In addition, the importance of the temporal dimension on intrusion detection, and the ability of this IDS to process it, are emphasized in this work.

Editorial: Hybrid learning machines

The concept of machine intelligence (MI) is complex, and thus many theories and definitions have emerged recently. Last few decades have seen a new era of machine intelligence focusing on the principles, theoretical aspects, and design methodology of algorithms gleaned from nature and biology. Examples are artificial neural networks inspired by mammalian neural systems, evolutionary computation inspired by natural selection in biology, simulated annealing inspired by thermodynamics principles, and swarm intelligence inspired by collective behavior of insects or microorganisms, etc. interacting locally with their environment causing coherent functional global patterns to emerge. These techniques have found their way in solving some real world problems in science, business, technology, and commerce. The integration of different learning and adaptation techniques, to overcome individual limitations and achieve synergetic effects through hybridization or fusion of these techniques, has in recent years contributed to a large number of new intelligent system designs. Despite the advances made, progress across the board has been moderate. One reason stems from the relatively slow pace at which work to understand biological intelligence has progressed. Another reason is rooted in the same inertia that has hampered the development of the research on intelligence previously—the reluctance to actively benefit from the achievements of the hybrid approaches. The 2nd International Workshop on Hybrid Artificial Intelligence Systems (HAIS 07–CAEPIA 2007) was held in conjunction with the Conference of the Spanish Association for Artificial Intelligence (CAEPIA) in Salamanca, Spain, from 12 to 16 November 2007. It was organized by the Biomedicine, Intelligent Systems and Educational Technology Group of the University of Salamanca. HAIS 2007 gathered individual researchers who see the need for synergy between various intelligent techniques. This special issue comprising 10 papers is focused on different hybrid learning approaches and its real world applications. Papers were selected on the basis of fundamental ideas/concepts rather than the thoroughness of techniques deployed. The papers are organized as follows. In the first paper, Gutiérrez et al. propose a hybrid neural network model using a possible combination of different transfer projection functions and kernel functions in the hidden layer of a feed-forward neural network. An evolutionary algorithm is adapted to this model and applied for learning the architecture, weights and node typology. Three different combined basis function models are proposed with all the different pairs that can be obtained. Combined functions using projection and kernel functions are found to be better than pure basis functions for the task of classification in several data sets.

Maximum and Minimum Likelihood Hebbian Learning for Exploratory Projection Pursuit

In this paper, we review an extension of the learning rules in a Principal Component Analysis network which has been derived to be optimal for a specific probability density function. We note that this probability density function is one of a family of pdfs and investigate the learning rules formed in order to be optimal for several members of this family. We show that, whereas we have previously (Lai et al., 2000; Fyfe and MacDonald, 2002) viewed the single member of the family as an extension of PCA, it is more appropriate to view the whole family of learning rules as methods of performing Exploratory Projection Pursuit. We illustrate this on both artificial and real data sets.

Intelligent Data Engineering and Automated Learning - IDEAL 2009

Learning and Information Processing.- Taking Advantage of Class-Specific Feature Selection.- Local Approximations.- SCIS: Combining Instance Selection Methods to Increase Their Effectiveness over a Wide Range of Domains.- Supervised Feature Extraction Using Hilbert-Schmidt Norms.- A Novel Estimation of the Regularization Parameter for ?-SVM.- Nearest Neighbor Classification by Relearning.- Integrating Rough Set and Genetic Algorithm for Negative Rule Extraction.- Development of a Conceptual Model for a Knowledge-Based System for the Design of Closed-Loop PID Controllers.- Lazy Classification Using an Optimized Instance-Based Learner.- Adaptive Fuzzy Logic Controller and Its Application in MEMS Mirror Actuation Feedback Control.- Detecting Computer Intrusions with Bayesian Networks.- Phase Load Balancing in the Secondary Distribution Network Using a Fuzzy Logic and a Combinatorial Optimization Based on the Newton Raphson.- Imperfect Pattern Recognition Using the Fuzzy Measure Theory.- K-Means Clustering Seeds Initialization Based on Centrality, Sparsity, and Isotropy.- Recurrence-Based Synchronization of Single Trials for EEG-Data Analysis.- FeedRank: A Semantic-Based Management System of Web Feeds.- An Autonomous Learning Algorithm of Resource Allocating Network.- Real-Time Nose Detection and Tracking Based on AdaBoost and Optical Flow Algorithms.- Hand Localization and Fingers Features Extraction: Application to Digit Recognition in Sign Language.- Interaction Detection in Aerodynamic Design Data.- Semi-supervised Outcome Prediction for a Type of Human Brain Tumour Using Partially Labeled MRS Information.- Optimizing Data Transformations for Classification Tasks.- The Minimum Redundancy - Maximum Relevance Approach to Building Sparse Support Vector Machines.- Discriminant Regression Analysis to Find Homogeneous Structures.- Learning from a Smarter Teacher.- STORM - A Novel Information Fusion and Cluster Interpretation Technique.- Discriminant Independent Component Analysis.- Information Preserving Empirical Mode Decomposition for Filtering Field Potentials.- Data Mining and Information Management.- A Heuristic Partial-Correlation-Based Algorithm for Causal Relationship Discovery on Continuous Data.- Clustering with XCS and Agglomerative Rule Merging.- Extended Cascaded Star Schema and ECOLAP Operations for Spatial Data Warehouse.- Block Clustering for Web Pages Categorization.- Framework for Decisional Business Modeling and Requirements Modeling in Data Mining Projects.- An AI Tool for the Petroleum Industry Based on Image Analysis and Hierarchical Clustering.- Quantitative Association Rules Applied to Climatological Time Series Forecasting.- Duplicate Candidate Elimination and Fast Support Calculation for Frequent Subgraph Mining.- Knowledge Extraction with Non-Negative Matrix Factorization for Text Classification.- Spherical Harmonics and Distance Transform for Image Representation and Retrieval.- Fourier Transform Based Spatial Outlier Mining.- Fuzzy Multi-Criteria Decision Making in Stereovision Matching for Fish-Eye Lenses in Forest Analysis.- Fuzzy Query Model for XML Documents.- Similarity-Binning Averaging: A Generalisation of Binning Calibration.- Compressed Disjunction-Free Pattern Representation versus Essential Pattern Representation.- Neuro-Informatics, Bio-Informatics and Bio-Inspired Models.- Combining Multiple Evolved Analog Circuits for Robust Evolvable Hardware.- Web Feed Clustering and Tagging Aggregator Using Topological Tree-Based Self-Organizing Maps.- A Hybrid Grouping Genetic Algorithm for the Multiple-Type Access Node Location Problem.- A Comparative Study of Stellar Spectra Analysis with Neural Networks in Transformed Domains.- Cascade-Connected ANN Structures for Indoor WLAN Positioning.- The Spatial Pheromone Signal for Ant Colony Optimisation.- Intrusion Detection in Sensor Networks Using Clustering and Immune Systems.- Novel Architecture for RNA Secondary Structure Prediction.- Nonlinear Dimensionality Reduction for Face Recognition.- A Framework for Pattern-Based Global Models.- A New Segmentation Approach in Structured Self-Organizing Maps for Image Retrieval.- GPU Implementation of the Multiple Back-Propagation Algorithm.- LDA Pre-processing for Classification: Class-Dependent Single Objective GA and Multi-objective GA Approaches.- Neural Network with Classification Based on Multiple Association Rule for Classifying Mammographic Data.- A Fuzzy Approach for Studying Combinatorial Regulatory Actions of Transcription Factors in Yeast.- Agents and Hybrid Systems.- The Winning Advantage: Using Opponent Models in Robot Soccer.- Talking Agents Design on the ICARO Framework.- A Rule-Based Multi-agent System for Local Traffic Management.- Requirements Engineering in the Development of Multi-Agent Systems: A Systematic Review.- Resources Oriented Search: A Strategy to Transfer Knowledge in the TRIZ-CBR Synergy.- Agent Negotiation Protocols in Time-Bounded Service Composition.- Writer Identification Using a Hybrid Method Combining Gabor Wavelet and Mesh Fractal Dimension.- Soft Computing Techniques in Data Mining.- Segmentation of Upwelling Regions in Sea Surface Temperature Images via Unsupervised Fuzzy Clustering.- Exploration of Bagging Ensembles Comprising Genetic Fuzzy Models to Assist with Real Estate Appraisals.- Implementation and Integration of Algorithms into the KEEL Data-Mining Software Tool.- A Niching Algorithm to Learn Discriminant Functions with Multi-Label Patterns.- Fuzzy Quantification-Based Linguistic Summaries in Data Cubes with Hierarchical Fuzzy Partition of Time Dimension.- A Soft Discretization Technique for Fuzzy Decision Trees Using Resampling.- Evolving Fuzzy Systems Based on the eTS Learning Algorithm for the Valuation of Residential Premises.- GFS-Based Analysis of Vague Databases in High Performance Athletics.- Recent Advances on Swarm-Based Computing.- The Vector Model of Artificial Physics Optimization Algorithm for Global Optimization Problems.- Electricity Consumption Simulation Based on Multi-agent System.- Using Preferences to Solve Student-Class Allocation Problem.- Nearest Neighbor Interaction PSO Based on Small-World Model.- Intelligent Computational Techniques in Medical Image Processing.- Classification Results of Artificial Neural Networks for Alzheimers Disease Detection.- An Automatic Segmentation and Reconstruction of Mandibular Structures from CT-Data.- Stent Graft Change Detection After Endovascular Abdominal Aortic Aneurysm Repair.- Segmentation of Abdominal Aortic Aneurysms in CT Images Using a Radial Model Approach.- Advances on Ensemble Learning and Information Fusion.- Interval-Valued Fuzzy Observations in Bayes Classifier.- Random Relevant and Non-redundant Feature Subspaces for Co-training.- Modification of Nested Hyperrectangle Exemplar as a Proposition of Information Fusion Method.- Financial and Business Engineering (Modeling and Applications).- Modelling Evaluation of Railway Reform Level Using Fuzzy Logic.- A Comparison of Market Structures with Near-Zero-Intelligence Traders.- Evaluating the Performance of Adapting Trading Strategies with Different Memory Lengths.- MIR Day 2009 - Burgos.- Improving the Language Active Learning with Multiagent Systems.- A Multi-agent System to Learn from Oceanic Satellite Image Data.- A Proposal for an Optimal Mutation Probability in an Evolutionary Model Based on Turing Machines.- Segmentation and Classification of Time-Series: Real Case Studies.- A Compendium of Heuristic Methods for Scheduling in Computational Grids.- Modeling of Network Computing Systems for Decision Tree Induction Tasks.- Atmospheric Pollution Analysis by Unsupervised Learning.- Improving Energy Efficiency in Buildings Using Machine Intelligence.- Analysis, Design and Implementation of a Multiagent System, to Extract Defining Contexts Based on a Linguistic Corpus in the Neurological Disease Domain.- Nature Inspired Models for Industrial Applications.- Applying Scatter Search to the Location Areas Problem.- Parameter Analysis for Differential Evolution with Pareto Tournaments in a Multiobjective Frequency Assignment Problem.- SOM-Based Selection of Monitored Consumers for Demand Prediction.- Multiagent Systems for Power System Topology Verification.

A soft computing method for detecting lifetime building thermal insulation failures

The detection of thermal insulation failures in buildings in operation responds to the challenge of improving building energy efficiency. This multidisciplinary study presents a novel four-step soft computing knowledge identification model called IKBIS to perform thermal insulation failure detection. It proposes the use of Exploratory Projection Pursuit methods to study the relation between input and output variables and data dimensionality reduction. It also applies system identification theory and neural networks for modeling the thermal dynamics of the building. Finally, the novel model is used to predict dynamic thermal biases, and two real cases of study as part of its empirical validation.

Soft computing models to identify typical meteorological days

Soft computing models are capable of identifying patterns that can characterize a ‘typical day’ in terms of its meteorological conditions. This multidisciplinary study examines data on six meteorological parameters gathered in a Spanish city. Data on these and other variables were collected for over 6 months, in 2007, from a pollution measurement station that forms part of a network of similar stations in the Spanish Autonomous Region of Castile– Leon. A comparison of the meteorological data allows relationships to be established between the meteorological variables and the days of the year. One of the main contributions of this study is the selection of appropriate data processing techniques, in order to identify typical days by analysing meteorological variables and aerosol pollutants. Two case studies are analysed in an attempt to identify a typical day in summer and in autumn.

Editorial: Editorial: New trends and applications on hybrid artificial intelligence systems

This Special Issue is an outgrowth of the HAIS’10, the 5th International Conference on Hybrid Artificial Intelligence Systems, which was held in San Sebastián, Spain, 23–25 June 2010. The HAIS conference series is devoted to the presentation of innovative techniques involving the hybridization of emerging and active topics in data mining and decision support systems, information fusion, evolutionary computation, visualization techniques, ensemble models, intelligent agent-based systems (complex systems), cognitive and reactive distributed AI systems, case base reasoning, nature-inspired smart hybrid systems, bioand neuro-informatics and their wide range of applications. It is dedicated to promote novel and advanced hybrid techniques as well as interdisciplinary applications and practice. HAIS’10 received over 269 submissions worldwide. After careful peerreview, only 132 papers were accepted for presentation at the conference and for inclusion in the proceedings, published as Springer’s Lecture Notes in Artificial Intelligence series. Authors of the most innovative papers within the scope of the NEUROCOMPUTING Journal were invited to submit their substantially extended and updated papers with additional original materials based on their most recent research findings. Each submitted paper was subsequently reviewed by 3–5 experts and leading researchers in the field. Finally eighteen papers passed the journal’s rigorous review process and were included in this Special Issue. They present an exclusive sample of the conference and its recent topics. In the area of artificial vision and image processing, Segovia et al. present a comparison between two methods for analyzing PET data in order to develop more accurate CAD systems for the diagnosis of Alzheimer’s disease. One of them is based on the Gaussian Mixture Model (GMM) and models the Regions Of Interest (ROIs) defined as differences between controls and AD subject. After GMM estimation using the EM algorithm, feature vectors are extracted for each image depending on the positions of the resulting Gaussians. The other method under study computes score vectors through a Partial Least Squares (PLS) algorithm based estimation and those vectors are used as features. Before extracting the score vectors, a binary mask based dimensional reduction of the input space is performed in order to remove low-intensity voxels. The validity of both methods is tested on the ADNI database by implementing several CAD systems with linear and nonlinear classifiers and comparing them with previous approaches such as VAF and PCA. The contribution of Chyzhyk et al. entitled ‘‘Hybrid Dendritic Computing with Kernel-LICA applied to Alzheimer’s disease detection in MRI’’ presents the issue of enhancing the generalization