New Advances in Intelligent Decision Technologies

Neural networks have been employed in many medical applications including breast cancer classification. Innovation in diagnostic features of tumours may play a central role in development of new treatment methods for earliest stage of breast cancer detection. Feature selection of neural network inputs is important for an accurate diagnosis application. Therefore, this study proposes elimination method for inputs feature selection of neural network to classify breast cancer cells. Morphological features were used as the inputs to several neural networks. The elimination method was employed on Hybrid Multilayer Perceptron (HMLP) network to investigate the diagnostic capability of features in combination and individually. Based on network performance resulted, the method was found practical for neural network inputs selection. Training the network with combination of dominant morphological features increased diagnosis capabilities and gave highest accuracy of 96%.

This publication presents a new concept of the intelligent analysis and interpretation of image-type data. The concept presented is based on a cognitive-semantic model of data learning and interpreting. The application of this model makes it possible to extract significant semantic information from the set of analysed data in order to understand it. It is also possible to execute the stage of reasoning based on the semantic contents of the analysed data. The proposed model for the cognitive analysis and interpretation of data is discussed using the example of a selected class of cognitive categorisation systems - UBIAS (Understanding Based Image Analysis Systems) analysing image-type data.

In the light of fast pace smart phone development, increasing numbers of applications are being developed to cater for portability. A real-time facial expression recognition application is develop that was tested in Windows Mobile environment. The underlying algorithm adopted in this work uses Boosting Naïve Bayesian (BNB) approach for recognition. We examine the structure of training data and the effect of attributes on the class probabilities through the use of Naïve Bayesian classifier (NBC). The experiments carried out show that we have achieved the important features of mobile application: speed and efficiency. This work is believed to be the first recorded initiative that de-ploys facial expression recognition into a mobile phone. It seeks to provide a launching point for a sound and portable mobile application that is capable of recognizing different facial expressions.

The popularity of multimedia on the Internet has created a heavy load on bandwidth. Consequently, content compression and bandwidth reduction have become significant topics recently. An appropriate codebook design is an essential and valuable principle for Vector Quantization (VQ). This investigation presents a new image compression method called INTSOM, which relies on Hierarchical Self-Organizing Map (HSOM) and adopts LBG for speeding up. For a two-layer neural network, INTSOM first employs LBG to determine the number of first layer neurons, and uses an estimation function to determine the number of second layer neurons dynamically. A modified SOM is then performed to compress each sub-map. Experimental results indicate that INTSOM has better overall capability (time-cost and quality) than LBG, SOM and HSOM.

While the importance of financial education increases in recent years, the technique for deepening an understanding of finance theory is needed. In this research, we analyize learning method of the finance theory about the investment project selection and capital structure determination using the business game technique. As a result of analysis, the participant understood the investment project selection method and interesting phenomena – an understanding progresses about the method of determining the capital structure which raises capital stock value – were seen. These results show the effectiveness of the business game technique to study of finance theory.

This work concerns the problem of modelling evolving prospective agent systems. Inasmuch a prospective agent [1] looks ahead a number of steps into the future, it is confronted with the problem of having several different possible courses of evolution, and therefore needs to be able to prefer amongst them to decide the best to follow as seen from its present state. First it needs a priori preferences for the generation of likely courses of evolution. Subsequently, this being one main contribution of this paper, based on the historical information as well as on a mixture of quantitative and qualitative a posteriori evaluation of its possible evolutions, we equip our agent with so-called evolution-level preferences mechanism, involving three distinct types of commitment. In addition, one other main contribution, to enable such a prospective agent to evolve, we provide a way for modelling its evolving knowledge base, including environment and course of evolution triggering of all active goals (desires), context-sensitive preferences and integrity constraints. We exhibit several examples to illustrate the proposed concepts.

We propose a new service composition method with the analytic hierarchy process and discuss its availability. The concept of context-aware services has been attracting attention as an approach to improving the usability of computer-mediated services. In ubiquitous computing environments, there are several means to provide services for users, and thus, to select an appropriate mean among them is a challenge. Our method for context-aware service composition determines service behaviors by context data. Through the implementation and examination of the method, we have found that the method can output reasonable results.

We show an improvement of our previous work, a formulation of the flexible widget layout (FWL) problem as a fuzzy constraint satisfaction problem (FCSP) and a method for solving it. The automation of widget layout is one of the most important challenges for the generation of graphical user interfaces (GUIs). In the field of model-based user interface design, widget layout is more complicated because a layout system needs to select widgets. FWL is the automatic GUI generation requiring (1) deciding which widgets are used and (2) completing the layout immediately. We formulate the desirability of selection as fuzzy constraints; thus, we can utilize existing techniques of FCSP without extending its framework.We divide the layout process into three phases, and realize the automatic layout in feasible time.

Humans know how to reason based on cause and effect, but cause and effect is not enough to draw conclusions due to the problem of imperfect information and uncertainty. To resolve these problems, humans reason combining causal models with probabilistic information. The theory that attempts to model both causality and probability is called probabilistic causation, better known as Causal Bayes Nets.

In this work we henceforth adopt a logic programming framework and methodology to model our functional description of Causal Bayes Nets, building on its many strengths and advantages to derive a consistent definition of its semantics. ACORDA is a declarative prospective logic programming which simulates human reasoning in multiple steps into the future. ACORDA itself is not equipped to deal with probabilistic theory. On the other hand, P-log is a declarative logic programming language that can be used to reason with probabilistic models. Integrated with P-log, ACORDA becomes ready to deal with uncertain problems that we face on a daily basis. We show how the integration between ACORDA and P-log has been accomplished, and we present cases of daily life examples that ACORDA can help people to reason about.

The major aim of library marketing system is to help the library and its patrons with providing useful information and various kinds of knowledge, which are extracted from the data that are collected by the library system. In this paper we lay heavy stress on the use of such information and knowledge for assisting decision making of the library and of its patrons. Furthermore our main concern is to investigate how much we can utilize the usage data of materials obtained in the intelligent bookshelves (IBSs), i.e. the bookshelves equipped with RFID antennas and their reader/writer controllers (R/Ws). In this paper we propose some analysis methods of these usage data, alone and combining with other library data, and demonstrate their potential importance for library marketing.

Designing optimal motion is critical in several applications for mobile robot from payload transport between two given states in a prescribed time such that a cost functional is minimized. This paper deals with the problem of path design of wheeled non-holonomic robots with flexible joints, based on Pontryagin’s minimum principle. The simplified case study of a Four Wheeled, two-link manipulator with joint elasticity is considered to study the method in generalized model. Nonlinear state and control constraints are treated without any simplifications or transforming them into sequences of systems with linear equations. By these means, the modeling of the complete optimal control problem and the accompanying boundary value problem is automated to a great extent. Performance of method is illustrated through the computer simulation.

Planning of robot trajectory is a very complex task that plays a crucial role in design and application of robots in task space. This paper is concerned with path planning of flexible robot arms for a given two-end-point task in point-to-point motion, based on indirect solution of optimal control problem. We employ the finite element method to modeling and deriving the dynamic equations of robot manipulator with flexible link, so in the presence of all nonlinear terms in dynamic equations open loop optimal control approach is a good candidate for generating the path that optimizes the end effector trajectory. Then the Hamiltonian function is formed and the necessary conditions for optimality are derived from the Pontryagin’s minimum principle. The obtained equations establish a two point boundary value problem which is solved by numerical techniques. Finally, simulations for a two-link planar manipulator with flexible links are carried out to investigate the efficiency of the presented method. The results illustrate the power and efficiency of the method to overcome the high nonlinearity nature of the problem.

Coalition stability is an important concept in coalition formation. One common assumption in many stability criteria in non-transferable utility games is that the preference of each agent is publicly known so that a coalition is said to be stable if there is no objections by any sub-group of agents according to the publicly known preferences. However, in many software agent applications, this assumption is not true. Instead, agents are modeled as individuals with private belief and decisions are made according to those beliefs instead of common knowledge. There are two types of uncertainty here. First, uncertainty in beliefs regarding the environment means that agents are also uncertain about their preference. Second, an agent’s actions can be influenced by his belief regarding other agents’ preferences. Such uncertainties have impacts on the coalition’s stability which is not reflected in the current stability criteria. In this paper, we extend the classic stability concept of the core by proposing new belief based stability criteria under uncertainty, and illustrate how the new concept can be used to analyze the stability of a new type of belief-based coalition formation game.

In order to decide on the course of action to take, one may need to check for side-effects of the possible available preferred actions. In the context of abduction in Logic Programs, abducible literals may represent actions and assumptions in the declarative rules used to represent our knowledge about the world. Besides finding out which alternative sets of actions achieve the desired goals, it may be of interest to identify which of those abductive solutions would also render

true

side-effect literals relevant for the decision making process at hand, and which would render those side-effects

false

. After collecting all the alternative abductive solutions for achieving the goals it is possible to identify which particular actions influence inspected side-effect literals’ truth-value.

To achieve this, we present the concept of Inspection Point in Abductive Logic Programs, and show how, by means of examples, one can employ it to investigate side-effects of interest (the

inspection

points

) in order to help evaluate and decide among abductive solutions. We show how this type of reasoning requires a new mechanism, not provided by others already available. We furthermore show how to implement this new mechanism it on top of an already existing abduction solving system — ABDUAL — in a way that can be adopted by other systems too.

Knowledge mining is the process of deriving new and useful knowledge from vast volumes of data and patterns previously discovered and stored as background knowledge. We propose a knowledge-mining system as a repertoire of tools for discovering strong and useful patterns. A pattern is strong if it represents frequently occurring relationships. Usefulness is achieved through constraints guided by users. To be able to derive strong and useful patterns from underlying data and background knowledge we consider employing the concept of higher-order logic as a major approach of our implementation. Higher-order logic can greatly reduce the burden of programmers as it is a very high level programming scheme suitable for the development of knowledge-intensive tasks. We have shown in this paper frequent pattern mining implemented with higher-order logic. The implementation is applied to mine breast cancer data. Our design of a logic-based knowledge-mining system is intended to support higher-order and constraint mining which is the next step of our research direction.

This paper describes a neural network-based classification tool that can be deployed for data-based decision support tasks. In particular, the Fuzzy ARTMAP (FAM) network is investigated, and a new pruning technique is proposed. The pruning technique is implemented successively to eliminate those rarely activated nodes in the category layer of FAM. Three data sets with different characteristics are used to analyze its effectiveness. In addition, a benchmark medical problem is used to evaluate its applicability as a decision support tool for medical diagnosis. From the experiment, the pruning technique is able to improve classification performances, as compared with those of to the original FAM network, as well as other machine learning methods. More importantly, the pruning technique yields more stable performances with fewer nodes, and results in a more parsimonious FAM network for undertaking data classification and decision support tasks.

The use of the Karhunen-Loève Transform (KLT) for the processing of the image primary color components gives as a result their decorrelation, which ensures the enhancement of such operations as: compression, color-based segmentation, etc. The basic problem is the high computational complexity of the KLT. In this paper is offered a simplified algorithm for the calculation of the KL color transform matrix. The presented approach is based on non-recursive approach for the color covariance matrix eigenvectors detection. The new algorithm surpasses the existing similar algorithms in its lower computational complexity, which is a prerequisite for fast color segmentation or for adaptive coding of color images aimed at real time applications.

Current research conducted at the Knowledge-Based Intelligent Information and Engineering Systems (KES) Centre aims to improve/develop the communication aspects of an agent-oriented architecture that enables agents to automatically adapt their functionality at runtime based on message flows. Rigid designtime constraints can be replaced by a flexible plug-and-play componentized capability. Intelligent Agents (IAs) must possess interoperability and capability to share knowledge and context in order to achieve their goal(s). A concept demonstrator is being developed, using a number of dynamic distributed environments, to show how interoperable Multi-Agent Systems (MASs) can improve data flow in a distributed environment. The agents in this MAS are equipped with a number of sensors that provide data from the environment, which is fused to produce knowledge. The fused information is fed into an inference engine which contains the Subject Mater Expert (SME) knowledge-based required to make decision(s) and/or change some course of action.

Experiments conducted by the Knowledge-Based Intelligent Information and Engineering Systems (KES) Centre use Java to gain its many advantages, especially in a distributed and dynamically scalable environment. Interoperability within and across ubiquitous computing operations has evolved to a level where

plug

‘

n

′

play

protocols that invoke common interfaces, provide the flexibility required for effective multi-lingual communications. One example includes: dynamic agent functionality within simulations that automatically adapt to incoming data and/or languages via scripts or messaging to achieve data management and inference. This has been shown using demonstrations at the Centre herein. Many aspects of the model involve web centric transactions, which involve data mining or the use of other types of Intelligent Decision Support System (IDSS). Section One of this paper provides an introduction, Section Two introduces the basic concepts of Decision Support System (DSS), Section Three discusses Intelligent Decision Support System (IDSS) enhancements, Section Four explains how agents use a multi-lingual dynamic environment,while Section Five highlights conclusions and future research direction.

The prime objective of intelligent agents in Multi-Agent Systems (MAS) is to act. An effective action results from a solid decision-making process. Decision-Support Systems (DSS) are used in MASs to assist in the development of a course of action for an individual or system goal. To ensure decision-making processes between agents remain objective and coherent, coordination model and cooperative problem-solving methodologies need to be implemented. Presently, coordination models have been developed as data-driven, process or control-driven, or hybrid models. Cooperative problem-solving methodologies have been designed to solely focus on allowing agents to share their knowledge which assists in achieving an individual goal or a course of actions. Although coordination and cooperation has been successfully implemented as separate frameworks within intelligent MASs, there is a significant limitation: cognitive modeling within each framework is limited or non-existent. This is a major obstacle within dynamic or unknown environments, as these cognitive environments heavily depend on precise information being made available to make well-informed and instantaneous decisions. This paper shows how the relationship between Belief-Desire-Intention (BDI) and Observe-Orient-Decide-Act (OODA) architectures, coordination and cooperation can promote decision-making processes in MASs. The linking of the decision-making process with coordination and cooperation can ameliorate their lack of cognitive processes. This enhancement is demonstrated by the decision support framework within the Agent Coordination and Cooperation Cognitive Model, or AC

3

M.

Safety onboard airborne platforms rests heavily on the way they are fixed. This fact includes repairs and testing, to reduce its down time. Maintenance practices using these components are achieved using generic and specific test equipment within the existing Maintenance Management System (MMS). This research paper reports the work performed to improve reliability and maintainability of Avionics Systems using an Intelligent Decision Support System (IDSS). In order to understand the shortcomings of the existing system, the prevalent practices and methodologies are researched. The paper reports the significant improvements made by integrating autonomous information sources as knowledge into an IDSS. Improvements are made by automating the existing data collection to create an expert system using intelligent agents. Data Mining techniques and intelligent agents are employed to create an expert system. Using feedback, the IDSS generates forecasts, alerts and warnings prior to system availability being compromised. If the data was stored electronically, Data Mining techniques and intelligent agents could be employed to create an expert system. Using feedback, an IDSS should generate forecasts or warnings prior to system availability being compromised. A Knowledge Base of all aspects of the logistics cycle is created as the system ages, to help make informed decisions about the platform, the Unit Under Test (UUT) or even the environment that supports it.

Automated guided vehicles (AGVs) are increasingly being used to transport goods or people. Navigation is a core issue in such a system. The AGVs use real-time control algorithms to reach their assigned destination autonomously. For reasons like scalability and flexibility, it is beneficial that the shuttles compute the necessary calculations decentrally. In this paper, we present such decentralized algorithms for conflict-free routing in a specific AGV system. Based on existing algorithms for deadlock handling in theory and routing in computer networks, we implemented three different sets of algorithms of varying sophistication in a logistics simulator. Evaluation reveals their functionality and relative performance.

Decision support systems are designed to assist a user by delivering targeted support for a specific decision problem. The application domain is an important consideration in developing the system, and the decision problem defines needed data, models and processing. Dynamic environments offer unique challenges to the system designer since the situation can change rapidly, requiring modification to the supporting technology. Modern intelligent techniques make it possible to envision a new type of adaptive decision support system (ADSS) that can perceive context and associate relevant models and data with the decision problem. In this paper we propose a framework for ADSS that utilizes intelligent methods to contextualize the decision problem and provide appropriate support to the user.

Landmines pose a significant problem in many countries around the world. Although technological systems such as metal detectors have been employed to combat these threats, many of these still require significant human interaction especially in the area of target and clutter discrimination. The aim of this research is to develop an automated decision making system for landmine detection. The initial stages of the research involves comparing various techniques for feature extraction to determine which methods provide the best representation for metal detector data to achieve improved target discrimination from background noise. This paper will focus on evaluating a technique utilizing the Continuous Wavelet Transform with false alarm rate and probability of detection used as performance measures.

The exploration of three-dimensional (3D) anthropometry scanning data along with other existing subject medical profiles using data mining techniques becomes an important research issue for medical decision support. This research attempts to construct a classification approach based on the hybrid of the case-based reasoning (CBR) and genetic algorithms (GAs) approach for hypertension detection using anthropometric body surface scanning data. The experiment showed that our proposed approach is able to improve the effectiveness of case matching of hypertension disease.

The concept of triggers has been around for more than two decades. Despite their diverse potential usages, trigger rules are difficult to define correctly and have to be carefully hand-coded by database programmers. We suggest an automatic way of trigger rule creation by the advanced technology of data mining. We propose a framework of trigger rule induction as well as a method for trigger conflict resolution. On trigger firing the problem may arise if several trigger rules are eligible for execution. We propose a conflict resolution scheme that incorporates derived knowledge as a major part of the trigger rule prioritization. By means of trigger scheduling, deterministic behavior of the trigger processing can be guaranteed. We demonstrate the utilization of our proposed method on enhancing medical database consistency.

This paper shows how an efficient and parallel algorithm for inference in Bayesian Networks (BNs) can be built and implemented combining sparse matrix factorization methods with variable elimination algorithms for BNs. This entails a complete separation between a first symbolic phase, and a second numerical phase.

In this paper, the approach of incorporating Hopfield neural networks (HNN) into ant colony systems (ACS) is proposed and studied. In the proposed approach (HNNACS), HNN is used to find a plausibly good solution, which is then used in ACS as the currently best tour for the offline pheromone trail update. The idea is to deposit additional pheromone to ACS to enhance the search efficiency. From simulation results, the search efficiency of HNNACS is better than other existing algorithms.

All over the world cardiovascular diseases are responsible for a great number of clinical problems that result in death. Arterial Hypertension is considered one of the main agents of such diseases and, for that reason, the correct detection of Hypertension symptoms is very important for the medical area. In this work we use the Paraconsistent Logic (PL) as a method of treatment of uncertainties to help health professionals that act in this medical area. With base in the concepts of the Paraconsistent Logic we created a Expert System for the help in the analysis and diagnosis in cardiovascular diseases. This computational system is called “Para

Hyper

Analyzer” and it uses Algorithms of the Paraconsistent Logic, a non-classical kind of logic, for the treatment of contradictory information. The algorithms are obtained through the interpretation of the representative Lattice of Paraconsistent Annotated Logic with annotation of two values (PAL2v), where the certainty degree (D

C

) and the contradiction degree (D

ct

) are calculated, according to the logical methodology. The PAL2v algorithms are called Paraconsistent Analysis Nodes (PAN). The PAN allows the analysis of propositions which are related to changes in the blood pressure, so that the patient can be classified in a certain risk group. The Para

Hyper

Analyzer studies this information and presents a result which shows the possibility of developing a serious cardiovascular disease.

This article presents a two level hierarchical forecasting model developed in a consulting project for a Brazilian magazine publishing company. The first level uses a VARMA model and considers econometric variables. The second level takes into account qualitative aspects of each publication issue, and is based on polynomial networks generated by Genetic Programming (GP).

This work presents some improvements regarding to the autonomous mobile robot Emmy based on Paraconsistent Annotated Evidential Logic E

τ

. A discussion on navigation system is presented.

A simple method for enhancing the output images of impulse noise filters for digital images is presented. The method is based on an edge detector and a simple 3-input 1-output neuro-fuzzy network. The internal parameters of the neuro-fuzzy network are adaptively optimized by training. The training is easily accomplished by using simple artificial images generated in a computer. The method can be used with any type of impulse noise filters since its operation is completely independent of the filter. The proposed method is applied to four representative impulse noise filters from the literature under different noise conditions and image properties. Results indicate that the proposed method may efficiently be used with any type of impulse noise filters to effectively reduce its distortion effects and enhance its output.

In this paper we present an algebraic version of the monadic system C

1

* of Da Costa [6] by using the concept of Curry Algebra [4]. The algebraic structure obtained is called Curry Algebra C

1

*. Some basic properties are discussed and presented.

To interpret the truth-value of future contingent events is of special importance for philosophy since the age of Aristotle. The subject is also interesting from the perspectives of linguistics and computer science. We compare two three-valued temporal logics for future contingents. We also suggest formalizing other types of useful three-valued temporal logics.

In this paper, in order to deal with plausible (deontic) reasoning in the framework of EVALPSN as well as defeasible deontic one, we introduce translation rules from Billington’s plausible theory into EVALPSN and show that the provability of the plausible theory can be preserved by the translation with a simple example.

Text classification aims to classify documents into categories or classes automatically based on their contents. While more and more textual information is available online, effective retrieval becomes difficult without good indexing and summarization of document contents. A rough set based reduct is a minimal subset of features, which has almost the same discernible power as the entire conditional features. Here, we propose a greedy algorithm to compute a set of rough set reducts which is followed by the k-nearest neighbor to classify documents. To improve the classification performance, reducts-kNN with confidence was developed. These proposed rough set reduct based classification methods for documents are compared by classification experiments. Experiments have been conducted on some benchmark datasets from the Reuters news data set.

A heterogeneous multi-agent system is a system that involves two or more agents that cooperate in order to accomplish a certain task. Genetic Network Programming (GNP) is a technique to automatically build a multi-agent system. In the past, the authors proposed the use of the Immune evolved Genetic Network Programming (IGNP) technique for the automatic construction of multi-agent systems. In this paper, the authors propose the use of Genetic Network Programming with Immune Adjustment Mechanism (GNPIAM) as a technique to automatically build a heterogeneous multi-agent system. In this study, the authors carry out experiments using tile world to evaluate the validity of the proposed method and compare the three techniques—GNP, IGNP, and GNPIAM.

In this paper we give an analysis of student friendship relation using class attendance records recorded by an attendance record system installed in Nagoya Institute of Technology (NIT). Our previous work[4] gives a prediction method of friendship relation among students using the data. This paper reveals further detailed analysis of friend and shows a possibility to have asymmetric relation among friendship from attendance records.

Particle filtering based motion tracking needs the extraction of the region of moving object to assign particles for moving object. To extract the moving object, a background subtraction is often used. However, it is difficult to extract the moving object when illumination changes during motion. Updating background image using RANSAC has been proposed to solve this problem, but it is still difficult for RANSAC to update the background image with high accuracy when many exception values are included in the data for updating background. In addition, another constraint includes such that the first background image is necessary for updating background image to extract the moving object. This paper proposes an extended new approach to update the background image with high accuracy using the data which excepts the exception values based on the tracking result with particle filtering. PSA (Pixel State Analysis) is further introduced to distribute particles before updating the first background, which can assign particles without preparing background image in advance.

This paper attempts to present a comprehensive survey of what is being done to automate the drowsiness ratings to be employed within a vehicle. The paper analyses the evidences for the usefulness of the measures currently used in drowsiness detection devices, which are not invasive and is based solely on eye activity. Their relationships with drowsiness and performance are described, and general problems and pitfalls associated with their practical use in passenger vehicles are identified. It also simulates a non-intrusive drowsiness detection system that is the core detection technique of several devices under review to understand how all the components of the system respond in real-time. A rating table to aid in automating the drowsiness rating in future is also included based upon analysis of drowsiness observed from recorded video.

We present a new set of rapid detection of background subtraction algorithms using codebooks to established Background Model (BG Model) and the concept of Color Model originally proposed by [6]. The proposed methods do not require prior learning, as in [6], and can create an instant BG Model detection and training with instant learning mechanism. Our proposed methods can also turn the latter coming but stationary foreground objects gradually as background, which are more adaptive to the actual environments. The proposed methods can also use instant learning to absorb sudden camera movements caused by the environments. We show that the proposed methods are effective and efficient in video surveillance applications.

One of the main difficulties in most modern Intrusion Detection Systems is the problem of massive alarms generated by the systems. The alarms may either be false alarms which are wrongly classified by a sensitive model, or duplicated alarms which may be issued by various intrusion detectors or be issued at different time for the same attack. We focus on learning-based alarm filtering system. The system takes alarms as the input which may include the alarms from several intrusion detectors, or the alarms issued in different time such as for multi-step attacks. The goal is to filter those alarms with high accuracy and enough representative capability so that the number of false alarms and duplicated alarms can be reduced and the efforts from alarm analysts can be significantly saved. To achieve that, we consider the causal correlation between relevant alarms in the temporal domain to re-label the alarm either to be a false alarm, a duplicated alarm, or a representative true alarm. To be more specific, recognizing the importance of causal correlation can also help us to find novel attacks. As another feature of our system, our system can deal with the frequent changes of network environment. The framework gives the judgment of attacks adaptively. An ensemble of classifiers is adopted for the purpose. Accordingly, we propose a system mainly consisting of two components: one is for alarm filtering to reduce the number of false alarms and duplicated alarms; and one is the ensemble-based adaptive learner which is capable of adapting to environment changes through automatic tuning given the expertise feedback. Two datasets are evaluated.

Outlier detection is an important issue in datamining and has been studied in different research areas. It can be used for detecting the small amount of deviated data. In this article, we use “Leave One Out” procedure to check each individual point the “with or without” effect on the variation of principal directions. Based on this idea, an over-sampling principal component analysis outlier detection method is proposed for emphasizing the influence of an abnormal instance (or an outlier). Except for identifying the suspicious outliers, we also design an on-line anomaly detection to detect the new arriving anomaly. In addition, we also study the quick updating of the principal directions for the effective computation and satisfying the on-line detecting demand. Numerical experiments show that our proposed method is effective in computation time and anomaly detection.

Interpretive structural modelling (ISM) is a useful method employed in decision making in industrial and systems engineering fields. Moreover, ISM often plays an important role in structuralising particular issues or problems related to medical issues. A small number of elements can be straightforwardly calculated using ISM, but it is difficult to structuralise the problem with a large number of elements using electronic computers in polynomial time. In the real world, medical decision support systems (MDSS) are basically composed of electronic computer-based systems. Therefore, in this paper, we show results on the basis of using a bio-inspired ISM that measures the efficiency of combining a computer-based decision support system towards the creation of a new MDSS, using an example of a rehabilitation centre selection problem.

The most traditional formative education of doctors is to have students to simulate diagnosis and treatment modes of their teachers, or carry handbooks such as “clinical summary” and “suit the remedy to the case for clinical illness” with them as the guideline. However, the temporary approach by finding answers in the books is not only time-consuming, but also affects the confidence of the patients to the doctor. This research aims at the above-said clinical practice in clinical emergency call and doctor formative education. We take the ophthalmology as an example; utilize the back-propagation algorithm of the artificial neural networks, to construct an “approximation decision model of medical record” for clinical diagnosis guideline. The doctor can input information such as chief complaint, other complaint and diagnosis etc. into the decision model and then correct ophthalmologic approximation medical records are outputted as the reference of diagnosis and treatment to improve the quality of medical treatment and medical care.

The clinical symptoms of metabolic disorders during neonatal period are often not apparent. If not treated early, irreversible damages such as mental retardation may occur, even death. Therefore, practicing newborn screening is essential, imperative to prevent neonatal from these damages. In the paper, we establish a newborn screening model that utilizes Support Vector Machines (SVM) techniques and enhancements to evaluate, interpret the Methylmalonic Acidemia (MMA) metabolic disorders. The model encompasses the Feature Selections, Grid Search, Cross Validations as well as multi model Voting Mechanism. In the model, the predicting accuracy, sensitivity and specificity of MMA can be improved dramatically. The model will be able to apply to other metabolic diseases as well.

Patients’ safety is the most essential, critical issue, however, errors can hardly prevent, especially for human faults. In order to reduce the errors caused by human, we construct Electronic Health Records (EHR) in the Health Information System (HIS) to facilitate patients’ safety and to improve the quality of medical care. During the medical care processing, all the tasks are based upon physicians’ orders. In National Taiwan University Hospital (NTUH), the Electronic Health Record committee proposed a localized standard of flows of orders. There are objectives of the standard: first, to enhance medical procedures and enforce hospital policies; secondly, to improve the quality of medical care; third, to collect sufficient, adequate data for EHR in the near future. Among the proposed procedures, NTUH decides to establish a Mobile Electronic Medication Administration Record (ME-MAR) System. According to researches, it indicates that medication errors are highly proportion to total medical faults. Therefore, we expect the ME-MAR system can reduce medication errors. In addition, we predict ME-MAR can assist nurses or healthcare practitioners to administer, manage medication properly.

Patient loyalty is a critical criterion for healthcare customer relationship management (CRM). An integrated framework with a case-based prediction model and a constraint-based optimization model is proposed to support the decision making of healthcare providers. This research first adopts a case-based prediction mechanism to forecast the possible loyalty level. We also proposes a constraint-based optimization approach as a subsequent mechanism to determine the optimum values of case features that may lead to the optimal patient loyalty. The potential use of this framework helps a decision maker allocate resources to increase the loyalty level for the given target patient segmentation.

RFID is an emerging technology intends to replace traditional barcode. Compared to barcode, RFID has the practical advantages of improving total product traceability and increasing accuracy. Especially healthcare service is life-critical, so that any careless mistakes may result in reversible loss. Due to the huge demand for automatic identification, many corporations and researchers devoted their efforts in RFID approaches for better achieving their goals. However, since there are various RFID solutions on the markets, it is critical to select the optimal solution to fit the desired scenarios. RFID deployment and implementation require not only technology concern, but also financial investment and user involvement. In this paper, we applied fuzzy TOPSIS to effectively evaluate suitable RFID solution providers. The proposed framework intends to help decision makers better evaluate the important factors affecting RFID implementation, forecasting the probability of a successful RFID project, as well as identifying the actions necessary before implementation. In addition, an experimental case study is presented to illustrate the application of the proposed approach.

We have been proposing a framework Rough Non-deterministic Information Analysis (

RNIA

), which handles rough sets based concepts in not only Deterministic Information Systems (

DISs

) but also Non-deterministic Information Systems (

NISs

). We have recently developed some algorithms and software tools for rule generation from

NISs

. Obtained rules characterize the tendencies in

NISs

, and they are often applied to decision making. However, if the condition parts in such rules are not satisfied, obtained rules are not applied to decision making. In this case, we need to examine each data in

NISs

, directly. In this paper, we add a question-answering with criterion values to

RNIA

. This addition enhances the aspect of decision making in

RNIA

.

This paper deals with cooperative games with

n

players and

r

alternatives which are called multi-alternative games with restricted choice situations. In these games, a value based on marginal contributions has been proposed. Many well-known values such as the Shapley value and the Banzhaf value are based on marginal contributions. On the other hand, some values such as the Deegan-Packel value are based on equally divided payoffs. Then, in this paper, we investigate a value based on equally divided payoffs for multi-alternative games with restricted choice situations.

In this paper, we described a relationship between topological spaces and neighborhood frames in modal logic. Then we introduce variable accessibility models for modal logic by, for each world, selecting a neighborhood adequate under a given context and/or time.

In this paper, we consider a heuristic method to partially calculate relative reducts with better evaluation by the evaluation criterion proposed by the authors. By using the average of certainty and coverage of decision rules constructed from each condition attribute, we introduce an evaluation criterion of condition attributes, and consider a heuristic method for calculating a relative reduct with better evaluation.

In this paper, we study rule induction based on the rough set theory. In the rough set theory, we induce minimal rules from a decision table, which is a data set composed of objects. Each object is described by condition attributes and classified by a decision attribute. When the decision attribute of a given decision table is ordinal, we may induce rules w.r.t. upward/downward unions of decision classes. This approach would be better in simplicity of obtained rules than inducing rules w.r.t. decision classes directly. However, because of independent applications of rule induction methods, inclusion relations among upward/downward unions in the conclusions of obtained rules are not inherited to the premises of those. This non-inheritance may debase the quality of obtained rules. In this paper, we propose two approaches to inherit the implication relations among the conclusions of obtained rules to the premises of those. Moreover, we propose an approach to classification unseen objects using rules w.r.t. upward/downward unions of decision classes. The performances of the proposed approaches are examined by numerical experiments.

The portfolio optimizations are generally to determine the proportion of funds in the portfolio consisting of the static assets. Then, it is hard to determine the proportion-weighted combination for the optimal portfolio consisting of the static large number of assets. In order to avoid this problem, we propose a Heuristic GA Method that optimizes the portfolio that consists of not only the given static assets but also the dynamically selected assets in this paper. In order to demonstrate the effectiveness of our method, we apply the method to creating an index fund based on correlation coefficients for the Tokyo Stock Exchange. This fund is one of the passively managed portfolios. The results show that our method works well for a dynamic index fund optimization.

This paper describes simulation studies in order to examine the validity of our previous predictions of the emergence of money using Doubly Structural Network Model (DSN model). DSN model consists of two levels of networks: the one of inner-agent model to represent their beliefs or knowledge about the world and the other of inter-agent model to represent a social network among agents. Using DSN model, we have explained how the concepts of money as a exchangeable media emerges through agent interaction. DSN Model is congenial to agent-based simulation. In this paper, using large scale intensive computer experiments, we investigate the bifurcation analysis derived from dynamics of DSN model. We also show new emergent phenomena on various types of social networks.

This article proposes methods to detect states of financial markets both comprehensively and with a high-resolution. In order to quantify trading patterns several mathematical methods are proposed based on frequencies of quotations/ transactions estimated from high-resolution data of financial markets. The empirical results (graphical network representation and quantification of states of market participants) for the foreign exchange market are shown. It is concluded that synchronous behavior associated with a large population of market participants may be a candidate of precursory signs leading to an environmental change.

We verify that Takayasu-Sato-Takayasu (TST) model satisfies not only Pareto’s law but also the detailed balance under Gibrat’s law, by using numerical simulation. We employ a tent-shaped function as multiplicative noise. We also numerically confirm that the reflection law is equivalent to the equation which gives the Pareto index

μ

in TST model. We extend the model modifying the stochastic coefficient under a Non-Gibrat’s law, and also numerically observe the detailed balance. The obtained pdf is power-law in the large scale region, and is the log-normal distribution in the middle scale one. We also study the dependence of Pareto index on the average of the additive noise.

We consider in this paper the effect of the player’s reputation implemented in a multi-agent model of iterated prisoner’s dilemma to develop cooperative networks. Our model assumes two separate strategies per agent to apply upon a cooperative partner and a defective partner. Starting from a randomly selected pair of strategies, (SD , SC) where SD and SC being the strategy on the defective partner and the cooperative partner, the agent autonomously learns a set of better strategies by imitating the better performing agent. Reputation is defined to be the rate of cooperative choice that the agent has chosen in the course of iteration. Each agent is given a fixed criterion on the minimum reputation to require upon the partner, so that this agent refuses to play the game with agents of bad reputation. We show by simulations that the model successfully develop cooperative networks of players by means of selecting the reputable partners as well as updating the strategies.

A group of intelligent agents exploring a time-frequency communications environment control a group of narrowband radio receivers. Multi-agent reinforcement learning is used in order to potentially increase the rate of processing wideband data and also to effectively co-ordinate the perceived narrowband information. TheMulti-Agent System (MAS) attempts to detect the presence of a single tone frequency agile radio transmission contained within the wider received band. Differing learning policies are investigated and the performance of the detector is considered.

A Bayesian estimation of a GARCH model is performed for US Dollar/Japanese Yen exchange rate by the Metropolis-Hastings algorithm with a proposal density given by the adaptive construction scheme. In the adaptive construction scheme the proposal density is assumed to take a form of a multivariate Student’s t-distribution and its parameters are evaluated by using the sampled data and updated adaptively during Markov Chain Monte Carlo simulations. We find that the autocorrelation times between the data sampled by the adaptive construction scheme are considerably reduced. We conclude that the adaptive construction scheme works efficiently for the Bayesian inference of the GARCH model.

We shed light on industrial structure of the economic system in Japan by combining visualization technique and community analysis. The production network consisting of submillion nodes (firms) and three million links (transactions) is visualized taking advantage of MD simulation technique. Also communities inherent in such a large-scale network is extracted through maximization of the modularity using both greedy (bottom-up) and bisection (top-down) algorithms; the bisection method works better. It is shown that nodes belonging to the same community are located close to each other in a visualization (three-dimensional) space.