Previous approaches for modeling Intrusion Detection System (IDS) have been on twofold: improving detection model(s) in terms of (i) feature selection of audit data through wrapper and filter methods and (ii) parameters optimization of detection model design, based on classification, clustering algorithms, etc. In this paper, we present three approaches to model IDS in the context of feature selection and parameters optimization: First, we present Fusion of Genetic Algorithm (GA) and Support Vector Machines (SVM) (FuGAS), which employs combinations of GA and SVM through genetic operation and it is capable of building an optimal detection model with only selected important features and optimal parameters value. Second, we present Correlation-based Hybrid Feature Selection (CoHyFS), which utilizes a filter method in conjunction of GA for feature selection in order to reduce long training time. Third, we present Simultaneous Intrinsic Model Identification (SIMI), which adopts Random Forest (RF) and shows better intrusion detection rates and feature selection results, along with no additional computational overheads. We show the experimental results and analysis of three approaches on KDD 1999 intrusion detection datasets.
Jaehun LEE Wooyong CHUNG Euntai KIM
A new structure learning approach for Bayesian networks (BNs) based on dual genetic algorithm (DGA) is proposed in this paper. An individual of the population is represented as a dual chromosome composed of two chromosomes. The first chromosome represents the ordering among the BN nodes and the second represents the conditional dependencies among the ordered BN nodes. It is rigorously shown that there is no BN structure that cannot be encoded by the proposed dual genetic encoding and the proposed encoding explores the entire solution space of the BN structures. In contrast with existing GA-based structure learning methods, the proposed method learns not only the topology of the BN nodes, but also the ordering among the BN nodes, thereby, exploring the wider solution space of a given problem than the existing method. The dual genetic operators are closed in the set of the admissible individuals. The proposed method is applied to real-world and benchmark applications, while its effectiveness is demonstrated through computer simulation.
Chen-Chien HSU Tsung-Chi LU Heng-Chou CHEN
In this paper, an evolutionary approach is proposed to obtain a discrete-time state-space interval model for uncertain continuous-time systems having interval uncertainties. Based on a worst-case analysis, the problem to derive the discrete interval model is first formulated as multiple mono-objective optimization problems for matrix-value functions associated with the discrete system matrices, and subsequently optimized via a proposed genetic algorithm (GA) to obtain the lower and upper bounds of the entries in the system matrices. To show the effectiveness of the proposed approach, roots clustering of the characteristic equation of the obtained discrete interval model is illustrated for comparison with those obtained via existing methods.
Mitsuyoshi KISHIHARA Isao OHTA Kuniyoshi YAMANE
This paper proposes a new type of compact waveguide directional coupler, which is constructed from two crossed E-plane rectangular waveguide with two metallic posts in the square junction and one metallic post at each port. The metallic posts in the square junction are set symmetrically along a diagonal line to obtain the directivity properties. The metallic post inserted at each input/output waveguide port can realize a matched state. Tight-coupling properties 0.79-6 dB are realized by optimizing the dimension of the junction and the positions/radii of the posts. The design results are verified by an em-simulator (Ansoft HFSS) and experiments.
Hiroyuki ISHIDA Tomokazu TAKAHASHI Ichiro IDE Yoshito MEKADA Hiroshi MURASE
We present a novel training method for recognizing traffic sign symbols. The symbol images captured by a car-mounted camera suffer from various forms of image degradation. To cope with degradations, similarly degraded images should be used as training data. Our method artificially generates such training data from original templates of traffic sign symbols. Degradation models and a GA-based algorithm that simulates actual captured images are established. The proposed method enables us to obtain training data of all categories without exhaustively collecting them. Experimental results show the effectiveness of the proposed method for traffic sign symbol recognition.
We propose a new genetic fuzzy discretization method with feature selection for the pattern classification problems. Traditional discretization methods categorize a continuous attribute into a number of bins. Because they are made on crisp discretization, there exists considerable information loss. Fuzzy discretization allows overlapping intervals and reflects linguistic classification. However, the number of intervals, the boundaries of intervals, and the degrees of overlapping are intractable to get optimized and a discretization process increases the total amount of data being transformed. We use a genetic algorithm with feature selection not only to optimize these parameters but also to reduce the amount of transformed data by filtering the unconcerned attributes. Experimental results showed considerable improvement on the classification accuracy over a crisp discretization and a typical fuzzy discretization with feature selection.
Akitsugu OHTSUKA Hirotsugu TANII Naotake KAMIURA Teijiro ISOKAWA Nobuyuki MATSUI
Data detection based on self organizing maps is presented for hematopoietic tumor patients. Learning data for the maps are generated from the screening data of examinees. The incomplete screening data without some item values is then supplemented by substituting averaged non-missing item values. In addition, redundant items, which are common to all the data and tend to have an unfavorable influence on data detection, are eliminated by a genetic algorithm and/or an immune algorithm. It is basically judged, by observing the label of a winner neuron in the map, whether the data presented to the map belongs to the class of hematopoietic tumors. Some experimental results are provided to show that the proposed methods achieve the high probability of correctly identifying examinees as hematopoietic tumor patients.
This paper presents an automated design of linear and non-linear differential analog circuits accelerated by reuse of genetic operations. The system first synthesizes circuits using pairs of simplified MOSFET model. During the evolutionary process, genetic operations that improve circuit characteristics are stored in a database and reused to effectively obtain a better circuit. Simplified elements in a generated circuit are replaced by MOSFETs and optimization of the transistor size is performed using an optimizer available in market if necessary. The capability of this method is demonstrated through experiments of synthesis of a differential voltage amplifier, a circuit having cube-law characteristic in differential mode and square-law characteristic in common-mode, and a dB-linear VGA (Variable Gain Amplifier). The results show the reuse of genetic operations accelerates the synthesis and success rate becomes 100%.
In recent years, network analysis has revealed that some real networks have the properties of small-world and/or scale-free networks. In this study, a simple Genetic Algorithm (GA) is regarded as a network where each node and each edge respectively represent a population and the possibility of the transition between two nodes. The characteristic path length (CPL), which is one of the most popular criteria in small-world networks, is derived analytically and shows how much the crossover operation affects the path length between two populations. As a result, the crossover operation is not so useful for shortening the CPL.
Wei SUN Yuanyuan ZHANG Yasushi INOGUCHI
Heterogeneous distributed computing environments are well suited to meet the fast increasing computational demands. Task scheduling is very important for a heterogeneous distributed system to satisfy the large computational demands of applications. The performance of a scheduler in a heterogeneous distributed system normally has something to do with the dynamic task flow, that is, the scheduler always suffers from the heterogeneity of task sizes and the variety of task arrivals. From the long-term viewpoint it is necessary and possible to improve the performance of the scheduler serving the dynamic task flow. In this paper we propose a task scheduling method including a scheduling strategy which adapts to the dynamic task flow and a genetic algorithm which can achieve the short completion time of a batch of tasks. The strategy and the genetic algorithm work with each other to enhance the scheduler's efficiency and performance. We simulated a task flow with enough tasks, the scheduler with our strategy and algorithm, and the schedulers with other strategies and algorithms. We also simulated a complex scenario including the variant arrival rate of tasks and the heterogeneous computational nodes. The simulation results show that our scheduler achieves much better scheduling results than the others, in terms of the average waiting time, the average response time, and the finish time of all tasks.
Hsien-Cheng TSENG Pei-Hsuan LEE Jung-Hua CHOU
An improved methodology, based on the genetic algorithm, is developed to design thermal-via structures and circuit parameters of advanced InGaP and InGaAs collector-up heterojunction bipolar transistors (C-up HBTs), which are promising miniature high-power amplifiers (HPAs) in cellular communication systems. Excellent simulated and measured results demonstrate the usefulness of this technique.
Rong-Long WANG Shinichi FUKUTA Jia-Hai WANG Kozo OKAZAKI
In this paper, we present a modified genetic algorithm for solving combinatorial optimization problems. The modified genetic algorithm in which crossover and mutation are performed conditionally instead of probabilistically has higher global and local search ability and is more easily applied to a problem than the conventional genetic algorithms. Three optimization problems are used to test the performances of the modified genetic algorithm. Experimental studies show that the modified genetic algorithm produces better results over the conventional one and other methods.
Ryosuke KUBOTA Keiichi HORIO Takeshi YAMAKAWA
In this paper, we propose a modified reproduction strategy of a Genetic Algorithm (GA) utilizing a Self-Organizing Map (SOM) with a novel updating rule of binary weight vectors based on a significance of elements of inputs. In this rule, an updating order of elements is decided by considering fitness values of individuals in a population. The SOM with the proposed updating rule can realize an effective reproduction.
Heng-Chou CHEN Oscal T.-C. CHEN
The probability associated with population fitness in a Genetic Algorithm (GA) is studied using the concept of average Euclidean distance. Based on the probability derived from population fitness, the GA can effectively terminate its evolution operations to mitigate the total computational load. Simulation results verify the feasibility of the derived probability used for the GA's termination strategy.
This paper presents an automated design of analog circuits starting with idealized elements. Our system first synthesizes circuits using idealized elements by a genetic algorithm (GA). GA evolves circuit topologies and transconductances of idealized elements to achieve the given specifications. The use of idealized elements effectively reduces search space and make the synthesis efficient. Second, idealized elements in a generated circuit are replaced by MOSFETs. Through the two processes, a circuit satisfying the given specifications can be obtained. The capability of this method was demonstrated through experiments of synthesis of a trans-impedance amplifier and a cubing circuit and benchmark tests. The results of the benchmark tests show the proposed CAD is more than 10 times faster than the CAD which does not use idealized elements.
Nonlinear modeling of complex irregular systems constitutes the essential part of many control and decision-making systems and fuzzy logic is one of the most effective algorithms to build such a nonlinear model. In this paper, a new approach to fuzzy modeling is proposed. The model considered herein is the well-known Sugeno-type fuzzy system. The fuzzy modeling algorithm suggested in this paper is composed of two phases: coarse tuning and fine tuning. In the first phase (coarse tuning), a successive clustering algorithm with the fuzzy validity measure (SCFVM) is proposed to find the number of the fuzzy rules and an initial fuzzy model. In the second phase (fine tuning), a moving genetic algorithm with partial encoding (MGAPE) is developed and used for optimized tuning of membership functions of the fuzzy model. Two computer simulation examples are provided to evaluate the performance of the proposed modeling approach and compare it with other modeling approaches.
Chen-Chien James HSU Chih-Yung YU Shih-Chi CHANG
Design of optimal controllers satisfying performance criteria of minimum tracking error and disturbance level for an interval system using a multi-objective evolutionary approach is proposed in this paper. Based on a worst-case design philosophy, the design problem is formulated as a minimax optimization problem, subsequently solved by a proposed two-phase multi-objective genetic algorithm (MOGA). By using two sets of interactive genetic algorithms where the first one determines the maximum (worst-case) cost function values for a given set of controller parameters while the other one minimizes the maximum cost function values passed from the first genetic algorithm, the proposed approach evolutionarily derives the optimal controllers for the interval system. To suitably assess chromosomes for their fitness in a population, root locations of the 32 generalized Kharitonov polynomials will be used to establish a constraints handling mechanism, based on which a fitness function can be constructed for effective evaluation of the chromosomes. Because of the time-consuming process that genetic algorithms generally exhibit, particularly the problem nature of minimax optimization, a parallel computation scheme for the evolutionary approach in the MATLAB-based working environment is also proposed to accelerate the design process.
We discuss a new design methodology of polynomial neural networks (PNN) in the framework of genetic algorithm (GA). The PNN is based on the ideas of group method of data handling (GMDH). Each node in the network is very flexible and can carry out polynomial type mapping between input and output variables. But the performances of PNN depend strongly on the number of input variables available to the model, the number of input variables, and the type (order) of the polynomials to each node. In this paper, GA is implemented to better use the optimal inputs and the order of polynomial in each node of PNN. The appropriate inputs and order are evolved accordingly and are tuned gradually throughout the GA iterations. We employ a binary coding for encoding key factors of the PNN into the chromosomes. The chromosomes are made of three sub-chromosomes which represent the order, number of inputs, and input candidates for modeling. To construct model by using significant approximation and generalization, we introduce the fitness function with a weighting factor. Comparisons with other modeling methods and conventional PNN show that the proposed design method offers encouraging advantages and better performance.
Takahiro YUKIZANE Shin-ya OHI Eiji MIYANO Hideo HIROSE
In difficult classification problems of the z-dimensional points into two groups giving 0-1 responses due to the messy data structure, we try to find the denser regions for the favorable customers of response 1, instead of finding the boundaries to separate the two groups. Such regions are called the bumps, and finding the boundaries of the bumps is called the bump hunting. The main objective of this paper is to find the largest region of the bumps under a specified ratio of the number of the points of response 1 to the total. Then, we may obtain a trade-off curve between the number of points of response 1 and the specified ratio. The decision tree method with the Gini's index will provide the simple-shaped boundaries for the bumps if the marginal density for response 1 shows a rather simple or monotonic shape. Since the computing time searching for the optimal trees will cost much because of the NP-hardness of the problem, some random search methods, e.g., the genetic algorithm adapted to the tree, are useful. Due to the existence of many local maxima unlike the ordinary genetic algorithm search results, the extreme-value statistics will be useful to estimate the global optimum number of captured points; this also guarantees the accuracy of the semi-optimal solution with the simple descriptive rules. This combined method of genetic algorithm search and extreme-value statistics use is new. We apply this method to some artificial messy data case which mimics the real customer database, showing a successful result. The reliability of the solution is discussed.
Mehdi EZOJI Karim FAEZ Hamidreza RASHIDY KANAN Saeed MOZAFFARI
Point pattern matching (PPM) arises in areas such as pattern recognition, digital video processing and computer vision. In this study, a novel Genetic Algorithm (GA) based method for matching affine-related point sets is described. Most common techniques for solving the PPM problem, consist in determining the correspondence between points localized spatially within two sets and then find the proper transformation parameters, using a set of equations. In this paper, we use this fact that the correspondence and transformation matrices are two unitary polar factors of Grammian matrices. We estimate one of these factors by the GA's population and then evaluate this estimation by computing an error function using another factor. This approach is an easily implemented one and because of using the GA in it, its computational complexity is lower than other known methods. Simulation results on synthetic and real point patterns with varying amount of noise, confirm that the algorithm is very effective.