Subtractive Clustering

This study introduces and evaluates novel concepts and techniques for automatic gender identification. Several approaches may be identified, including the utilization of a multilayer perceptron neural network in conjunction with the adaptive neuro-fuzzy inference system algorithm and genetic algorithms to enhance the optimization of network weights. Additionally, the application of neural networks for gender recognition and their integration with the fuzzy C-means method is also noteworthy. The most optimal outcome was achieved through the integration of the ANFIS network with the Fuzzy C-Means algorithm. Furthermore, alternative approaches proved to be highly efficacious. The highest achieved accuracy was observed for the Texas Instruments/Massachusetts Institute of technology dataset (97.5%) and for the Oregon Graduate Institute dataset (96.31%). The demonstrated high accuracy achieved in the analysis of Oregon Graduate Institute data, characterized by its multilingual phone data and low signal-to-noise ratio, indicates the robustness of the suggested approaches in handling variations in speaker language and the suboptimal quality of speech data. Furthermore, employing the genetic neural network methodology facilitated the development of a high-speed network capable of attaining comparable levels of accuracy to the multilayer perceptron neural network, albeit with a significantly reduced number of neurons in the middle layer, expressly limited to three.

This paper proposes HNF approach for the applications of AGC of a hydro-thermal and Thermal-thermal systems in a power system deregulated environment. It makes an attempt to provide a new practical AGC model with Generation Rate Constraints to fulfill the needs of modern restructured interconnected systems in a Deregulated Environment.

The success of the classification method is highly dependent on how to specify initial data as the initial prototype, dissimilarity functions that we used and the presence of outliers among the data. To overcome these obstacles, in this paper we present Fuzzy Kernel k-Medoids (FKkM) algorithm that we claim to be robust against outliers, invariant under translation and data transformation, as the combined development of Fuzzy LVQ, Fuzzy k-Medoids and Kernel Function. Based on the experiments, it provides a better accuracy than Support Vector Machines, Kernel Fisher Discriminant and RBF Neural Network for multiclass multidimensional data classification.

This paper proposes evaluation of Power System Restoration Indices (PSRI) based on the Automatic Generation Control (AGC) assessment of interconnected power system in a deregulated environment. This paper deals with the procedure involved in the AGC of a two-area reheat power system having a coordinated control action between Hydrogen Energy Storage (HES) unit and Unified Power Flow Controller (UPFC) for controlling the network performance in a very fast manner and to improve the power transfer limits in order to have a better restoration. In addition to that a new Proportional-Double Integral (PI 2 ) controller is designed and implemented in the AGC loop and the controller parameters are optimized through Bacterial Foraging Optimization (BFO) algorithm. Simulation results reveal that the proposed PI 2 controller ensures good stability during load variations, excellent transient and dynamic responses when compared with the system comprising PI controller. Moreover the AGC loop with ...

The representative load curves (RLCs) are necessary for utilities in tariffication policy. From the load curves collected in the activity time of a tariff, one representative load curve will be built. The easy way to estimate the impact of a tariff is to analyze some indicators of the RLC. The Fuzzy K-Means (FKM) is utilized in this work to determine RLCs. The compromising of the Cluster validity indexes and determining the suitable weighting exponent m are considered to find out the final clusters and their RLCs. The case study for one utility in the South of Vietnam is carried out to show the impacts of the current Time of Use (TOU) tariff.

This paper presents clustering techniques (K-means, Fuzzy K-means, Subtractive) applied on specific databases (Flower Classification and Mackey-Glass time series) , to automatically process large volumes of raw data, to identify the most relevant and significative patterns in pattern recognition, to extract production rules using Mamdani and Takagi-Sugeno-Kang fuzzy logic inference system types.

Clustering is one of the phenomenal process towards information retrieval and knowledge discovery. Cluster optimality is still a questionable factor for current benchmarking clustering strategies. In particular document clustering is most sensible towards information retrieval and knowledge discovery, which is due to the curse of high volume and high dimensionality observed in recent times. In order to this many of document clustering models have been devised in recent times, but all of these models are questionable either the case of cluster optimality, process time complexity or adoptability. Henceforth, here we devised a deep machine learning approach called incremental evolutionary genetic algorithm based optimal document clustering (ODC) process. The experiments were done on documents dataset with curse of high dimensionality and volume. The results obtained from the experiments observed to be remarkably optimistic towards document clustering and also evincing the linearity in ...

This paper presents the application of Adaptive Network Based Fuzzy Inference System ANFIS on speech recognition. The primary tasks of fuzzy modeling are structure identification and parameter optimization, the former determines the numbers of membership functions and fuzzy if-then rules while the latter identifies a feasible set of parameters under the given structure. However, the increase of input dimension, rule numbers will have an exponential growth and there will cause problem of "rule disaster". Thus, determination of an appropriate structure becomes an important issue where subtractive clustering is applied to define an optimal initial structure and obtain small number of rules. The appropriate learning algorithm is performed on TIMIT speech database supervised type, a pre-processing of the acoustic signal and extracting the coefficients MFCCs parameters relevant to the recognition system. Finally, hybrid learning combines the gradient decent and least square estimation LSE of parameters network. The results obtained show the effectiveness of the method in terms of recognition rate and number of fuzzy rules generated.

This paper explains how the commonly occurring DOS and Brute Force attacks on computer networks can be efficiently detected and network performance improved, which reduces costs and time. Therefore, network administrators attempt to instantly diagnose any network issues. The experimental work used the SNMP-MIB parameter datasets, which are collected via a specialised MIB dataset consisting of seven types of attack as noted in section three. To resolves such issues, this researched carried out several important contributions which are related to fault management concerns in computer network systems. A central task in the detection of the attacks relies on MIB feature behaviours using the suggested SFCM method. It was concluded that the DOS and Brute Force fault detection results for three different clustering methods demonstrated that the proposed SFCM detected every data point in the related group. Consequently, the FPC approached 1.0, its highest record, and an improved performance solution better than the EM methods and K-means are based on SNMP-MIB variables.

The success of the classification method is highly dependent on how to specify initial data as the initial prototype, dissimilarity functions that we used and the presence of outliers among the data. To overcome these obstacles, in this paper we present Fuzzy Kernel k-Medoids (FKkM) algorithm that we claim to be robust against outliers, invariant under translation and data transformation, as the combined development of Fuzzy LVQ, Fuzzy k-Medoids and Kernel Function. Based on the experiments, it provides a better accuracy than Support Vector Machines, Kernel Fisher Discriminant and RBF Neural Network for multiclass multidimensional data classification.

Background and Objectives: In this paper, a new version of the particle swarm optimization (PSO) algorithm using a linear ranking function is proposed for clustering uncertain data. In the proposed Uncertain Particle Swarm Clustering method, called UPSC method, triangular fuzzy numbers (TFNs) are used to represent uncertain data. Triangular fuzzy numbers are a good type of fuzzy numbers and have many applications in the real world. Methods: In the UPSC method input data are fuzzy numbers. Therefore, to upgrade the standard version of PSO, calculating the distance between the fuzzy numbers is necessary. For this purpose, a linear ranking function is applied in the fitness function of the PSO algorithm to describe the distance between fuzzy vectors. Results: The performance of the UPSC is tested on six artificial and nine benchmark datasets. The features of these datasets are represented by TFNs. Conclusion: The experimental results on fuzzy artificial datasets show that the proposed clustering method (UPSC) can cluster fuzzy datasets like or superior to other standard uncertain data clustering methods such as Uncertain K-Means Clustering (UK-means) and Uncertain K-Medoids Clustering (UK-medoids) algorithms. Also, the experimental results on fuzzy benchmark datasets demonstrate that in all datasets except Libras, the UPSC method provides better results in accuracy when compared to other methods. For example, in iris data, the clustering accuracy has increased by 2.67% compared to the UK-means method. In the case of wine data, the accuracy increased with the UPSC method is 1.69%. As another example, it can be said that the increase in accuracy for abalone data was 4%. Comparing the results with the rand index (RI) also shows the superiority of the proposed clustering method.

Background and Objectives: In this paper, a new version of the particle swarm optimization (PSO) algorithm using a linear ranking function is proposed for clustering uncertain data. In the proposed Uncertain Particle Swarm Clustering method, called UPSC method, triangular fuzzy numbers (TFNs) are used to represent uncertain data. Triangular fuzzy numbers are a good type of fuzzy numbers and have many applications in the real world. Methods: In the UPSC method input data are fuzzy numbers. Therefore, to upgrade the standard version of PSO, calculating the distance between the fuzzy numbers is necessary. For this purpose, a linear ranking function is applied in the fitness function of the PSO algorithm to describe the distance between fuzzy vectors. Results: The performance of the UPSC is tested on six artificial and nine benchmark datasets. The features of these datasets are represented by TFNs. Conclusion: The experimental results on fuzzy artificial datasets show that the proposed clustering method (UPSC) can cluster fuzzy datasets like or superior to other standard uncertain data clustering methods such as Uncertain K-Means Clustering (UK-means) and Uncertain K-Medoids Clustering (UK-medoids) algorithms. Also, the experimental results on fuzzy benchmark datasets demonstrate that in all datasets except Libras, the UPSC method provides better results in accuracy when compared to other methods. For example, in iris data, the clustering accuracy has increased by 2.67% compared to the UK-means method. In the case of wine data, the accuracy increased with the UPSC method is 1.69%. As another example, it can be said that the increase in accuracy for abalone data was 4%. Comparing the results with the rand index (RI) also shows the superiority of the proposed clustering method.

Fast methods for estimating voltage stability security limits are crucial in modern energy management systems. In this paper, a method to build a fuzzy inference system (FIS) is developed in order to estimate the loading margin. The main goal is to overcome the disadvantages of conventional methods and to apply this methodology in a real time operation environment. First, some voltage stability indices and variables are presented as candidate inputs to the FIS. Subtractive clustering is used to construct the initial FIS models, and adaptive neuro fuzzy inference systems allow tuning them so that it is possible to obtain better loading margin estimates. Extensive simulations were carried out in order to build data sets that take into account a quasi-random load direction, as well as information regarding base case and contingency situations, including branch, generator, and shunt single outages. Results are provided for the IEEE 30, 118, and 300 bus test systems.

In this article, a stochastic optimization technique called fuzzy particle swarm optimization (FPSO) is presented to determine an optimum set of microstrip antenna arrays excitation weights (amplitude and phase), the use of the fuzzy controller allows to dynamically adjust its parameters such as, the inertia weight and acceleration coefficients in order to produce an optimal pattern of the antenna array able to approach a desired pattern. Simulation results are proposed to compare with published results to verify the effectiveness of the suggest method for both linear and planar array.

A human face detection method for color images is presented in this paper, which is pose, size and position independent, and has the priority of classifying detected faces in three groups: frontal, near frontal and profile, according to their pose. This system is a fuzzy rule base one, optimized by genetic algorithm. In the first stage, skin color regions are selected in the input image. Within each skin area, lip pixels and ear texture are searched, and applied as features to identify face candidates in the skin regions. Summarizing all obtained information along by skin region shape and lip area position relative to skin area, four inputs are computed for fuzzy inference system, and face areas as well as their poses would be introduced. The proposed method is tried on various databases, including HHI, Champion, Caltech, Bao and IMM databases. Achieved results show a remarkable detection rates compared to other methods, for various face poses. 96.8%, 95.3% and 87.8% correct detecti...

This research applies Genetic Algorithm to find the initial cluster centers and the centers of this cluster will be used as an input for the K-Means method. This method yield a more optimal performance compared to the conventional K-Means method since the centers point is optimized with Genetic Algorithm. Unfortunately, k-means is extremely sensitive to the initial choice of centers, and a poor choice of centers may lead to a local optimum that is quite inferior to the global optimum. The proposed research initial clustering and assigns at least one seed point to each cluster. During the second step, the seed-points are adjusted to minimize the cost-function. The algorithm automatically penalizes any possible winning chances for all rival seed-points in subsequent iterations base on cost-function to reaches global minimum. This method will perform optimization centers point K-means clustering using Genetic Algorithms. The Genetic Algorithms optimize centers point of the cluster more...

It is hypothesized that a key characteristic of electrocardiogram (ECG) signal is its nonlinear dynamic behaviour and that the nonlinear component changes more significantly between normal and arrhythmia conditions than the linear component. The usual statistical descriptors used in RR (R to R) interval analysis do not capture the nonlinear disposition of RR interval variability. In this paper we explore a novel approach to extract the features from nonlinear component of the RR interval signal using Teager energy operator (TEO). The key feature of Teager energy is that it models the energy of the source that generated the signal rather than the energy of the signal itself. Hence any deviations in regular rhythmic activity of the heart get reflected in the Teager energy function. The classification evaluated on MIT-BIH database, with RR interval and mean of Teager energy computed over RR interval as features, exhibits an average accuracy that exceeds 99.79%.

ECG (Electrocardiogram) performs classification using a machine learning model for processing different features in the ECG signal. The electrical activity of the heart is computed with the ECG signal with machine learning library. The key issue in the handling of ECG signals is an estimation of irregularities to evaluate the health status of patients. The ECG signal evaluate the impulse waveform for the specialized tissues in the cardiac heart diseases. However, the ECG signal comprises of the different difficulties associated with waveform estimation to derive certain features. Through machine learning (ML) model the input features are computed with input ECG signals. In this paper, proposed a Noise QRS Feature to evaluate the features in the ECG signals for the effective classification. The Noise QRS Feature model computes the ECG signal features of the waveform sequences.  Initially, the signal is pre-processed with the Finite Impulse response (FIR) filter for the analysis of EC...

Recent advances in the nanorobots design and implementation field considers requirements that serve medical applications according to one of two scenarios: drug delivery and discovery tasks. These scenarios are based on detecting threshold concentration of specific chemicals emitted by cells or searching particular places in the human body and acting upon findings. These scenarios emphasize local self-coordination but lack the global coordination over long distances to cooperatively accomplish group task. Considering the scenario of launching a swarm of nanorobot in blood artery for the purpose of cutting cholesterol plaques, the proposed model identifies flocking based coordination between nanorobots in the swarm. The proposed model includes both decentralized and centralized coordination for possessing both the local information within the swarm and global information for interactive task assignment by the medical specialists. In addition the medical specialist can interactively cancel task, reassign new task whilst monitoring task accomplishment. The task is simply searching Cholesterol threshold size in a specified location in blood artery. The experimentation results has shown to be efficient as it overcomes coverage based particle swarm optimization problem.

Pap smear screening is the most successful attempt of medical science and practice for the early detection of cervical cancer. Manual analysis of the cervical cells is time consuming, laborious and error prone. This paper presents a Neural Network (NN) based system for classifying cervical cells as normal, low-grade squamous intra-epithelial lesion (LSIL) and high-grade squamous intra-epithelial lesion (HSIL). The system consists of three stages. In the first stage, cervical cells are segmented using the Adaptive Fuzzy Moving K-means (AFMKM) clustering algorithm. In the second stage, the feature extraction process is performed. In the third stage, the extracted data is classified using Fuzzy Min-Max (FMM) NN. The empirical results show that the proposed method can achieve acceptable results.

Background/Objectives: The main objective of this research is to design Deep Learning (DL) architecture to classify an electrocardiogram (ECG) signal into normal sinus rhythm (NSR), premature ventricular contraction (PVC), atrial premature contraction (APC) or right/left bundle branch block (RBBB/LBBB) arrhythmia by empirically optimizing the numbers of hidden layers, the number of neurons in each hidden layer and the number of neurons in input layer in DL model.Methods/Statistical analysis: For our experimental simulations, PhysioBank-MIT/BIH annotated ECG database was considered to classify heart beats into abnormal rhythms (PVC, APC, RBBB, LBBB) or normal sinus. The performance of classifying ECG beats by the proposed DL architecture was evaluated by computing the overall accuracy of classifying NSR or four different arrhythmias.Findings: Base on testing MIT/BIH arrhythmia database, the proposed DL model can classify the heart rhythm into one of NSR, PVC, APC, RBBB or LBBB beat w...

This research applies Genetic Algorithm to find the initial cluster centers and the centers of this cluster will be used as an input for the K-Means method. This method yield a more optimal performance compared to the conventional K-Means method since the centers point is optimized with Genetic Algorithm. Unfortunately, k-means is extremely sensitive to the initial choice of centers, and a poor choice of centers may lead to a local optimum that is quite inferior to the global optimum. The proposed research initial clustering and assigns at least one seed point to each cluster. During the second step, the seed-points are adjusted to minimize the cost-function. The algorithm automatically penalizes any possible winning chances for all rival seed-points in subsequent iterations base on cost-function to reaches global minimum. This method will perform optimization centers point K-means clustering using Genetic Algorithms. The Genetic Algorithms optimize centers point of the cluster more...

It is well known that the kingdom of Saudi Arabia is a vast natural potential for developing solar energy, there so solar power generation is growing rapidly. Solar energy depends on different weather and meteorological factors. Moreover, solar radiation variations throughout the year are considered an obstacle for predicting the solar energy. There so it is needed to develop predictive model that gives a precise estimation of the solar energy. The paper proposes effective model to give a reliable prediction of the global horizontal irradiance for the next day in Hail city. Estimating global horizontal irradiance has its influence on different outcomes for solar energy generation companies. It helps in reducing electricity production costs. The developed model applies subtractive clustering algorithm for generating initial model. Afterwards, it applies adaptive neuro-fuzzy inference system for tuning the parameters of the membership functions of the generated model. The simulation r...

Automatic recognition of cardiac arrhythmias is important for diagnosis of cardiac abnormalities. Detection of various abnormalities in the heart to identify various heart diseases can be made through an Adaptive Neuro-Fuzzy Inference System (ANFIS) preprocessed by subtractive clustering. Some recordings of the MIT-BIH database has been used for training and testing our neural network based classifier. Six types of heartbeats are classified: normal sinus rhythm, premature ventricular contraction (PVC), atrial premature contraction (APC), left bundle branch block (LBBB), right bundle branch block (RBBB), and paced beats. Our goal is to detect important characteristics of an ECG signal to determine if the heartbeat is normal or irregular. Therefore, in this paper, an expert system for Electrocardiogram (ECG) classification is analyzed. DWT is used in preprocessing for filtering ECG recordings, and extraction of some features performs the classification task. KeywordsANFIS, Arrhythmias...

Clustering is a process of partitioning similar data into groups. For this, number of clustering algorithms have been proposed in literature. Some of them can also be used for the generation of Fuzzy Models. In this work, Sugeno fuzzy models being generated by Subtractive and FCM clustering have been discussed. Experiments have been performed on real datasets to compare the Subtractive and FCM clustering. Further, the effect of increase in the radius size is analyzed in Subtractive clustering. The average absolute error and root mean square error is also found out when using FCM clustering and Subtractive clustering with different values of radius.

The main objective of the paper is to build a prediction system to predict the future occurrence of an event. Fuzzy logic, among the various available Artificial Intelligence techniques, emerges as an advantageous technique in predicting future events. Subjective and Objective modeling are two types of fuzzy modeling. Objective type fuzzy modeling is used to build the prediction system. It is a combination of a clustering algorithm and fuzzy system identification which proves effective in improving the efficiency of the prediction. To train the prediction system, historical data is obtained from the web. Data specific to the desired application is obtained and is recorded. This recorded information is subjectively reasoned to develop containing only the necessary inputs to the prediction system. The subtractive clustering algorithm is used for its computational advantages and fuzzy rules are formed using system identification technique. Stock markets are excellent examples where thi...

The Centroid-based clustering is an NP-hard optimization problem and the common approach is to search for cluster centers only for approximate solutions. In this paper we have proposed swarm intelligence based nature-inspired center-based clustering method using PSO optimization. Proposed PSO clustering method is capable to search best cluster with maximum fitness using social-only and cognition-only model, such that the square distances from the cluster are minimized. In this article, how PSO based clustering can be used to get N number of cluster specified by the user in a dataset is demonstrated. Our suggested method has been tested with artificial dataset and several datasets from UCI Machine learning repository. Effectiveness and usefulness of the proposed method is shown by comparing fitness of this method with Kmeans and Fuzzy c-means technique. For better comparative result, we have ended up with comparison of proposed clustering model with subtractive clustering (extension ...

Computer networks have experienced an explosive growth over the past few years and have become the targets for hackers and intruders. An intrusion detection system's main goal is to classify activities of a system into two major categories: normal activity and suspicious or intrusive activity. The objective of this paper is to expose ANFIS as a neuro-fuzzy classifier to detect intrusions in computer networks. Our experiments and evaluations were performed with the KDD Cup 99 intrusion detection dataset which is a version of the 1998 DARPA intrusion detection evaluation dataset prepared and managed by MIT Lincoln Laboratories. This paper shows that our proposed method can be effective in detecting various intrusions.

The statistical analysis of accident is conceded out periodically at grave locations or road stretch which will help to arrive at suitable measures to effectively decrease accident rates. It is the measure (or estimates) of the number and severity of accident. These statistics reports are to be maintained zone-wise. Accident prone stretches of various roads may be assessed by finding the accident density per meter of the road. The places of accidents are marked on the map and the points of their clustering (BLACK SPOT) are determined. With the help of statistical study of accident occurrence at a particular road or location or zone of study for a long period of time it is realizable to predict with logical accuracy the probability of accident occurrence per day or relative safety of different classes of road user in that location. The interpretation of the statistical data is very important to provide insight to the problem.

Magnetic resonance imaging (MRI) brain tumor segmen tation is a challenging tasks which include the detection task of tumor from images. In general, t his process is done manually by experts in medical images field which is always unclear, because the similarity between normal and abnormal tissues. The present study proposes a new clustering approach ba sed on the hybridization of firefly algorithm (FA) and fuzzy c-means algorithm(FCM) called (FFCM) to segme nt MRI brain images. this approach use the capability of firefly search to find optimal initia l cluster centres for the FCM and thus improve (MRI ) brain tumor segmentation. The proposed approach was evalu ated by applying it to a magnetic resonance imaging (MRI) brain segmentation problem using a simulated brain data set of McGill University and real MRI images from Internet Brain Segmentation Repository benchmark data sets. The cluster validity index (Rm) was used as a fitness function to determine the bes t solutions obtaine...

Software is the inseparable part of today’s human life. Each and every gadget that we use is dependent on some or other kind of software. Component based software engineering (CBSE) has provided an effective software development paradigm which allows selection of domain specific components from component repository and assemble them into a modular and scalable application. The reliability of software and its components depends on the amount of effective testing carried on it during its development life cycle. We cannot deny the fact that exhaustive testing is not possible. Selection of appropriate test suite is a combinatorial problem. Soft computing provides a promising solution for the same. Emergence of artificial intelligence over years has added fuel to nature inspired testing techniques. This paper is a comparative study of various soft computing approaches inspired by nature for reusable software components such as artificial neural network, genetic algorithm, fuzzy logic and...

This paper presents clustering techniques (K-means, Fuzzy K-means, Subtractive) applied on specific databases (Flower Classification and Mackey-Glass time series) , to automatically process large volumes of raw data, to identify the most relevant and significative patterns in pattern recognition, to extract production rules using Mamdani and Takagi-Sugeno-Kang fuzzy logic inference system types.

The Centroid-based clustering is an NP-hard optimization problem and the common approach is to search for cluster centers only for approximate solutions. In this paper we have proposed swarm intelligence based nature-inspired center-based clustering method using PSO optimization. Proposed PSO clustering method is capable to search best cluster with maximum fitness using social-only and cognition-only model, such that the square distances from the cluster are minimized. In this article, how PSO based clustering can be used to get N number of cluster specified by the user in a dataset is demonstrated. Our suggested method has been tested with artificial dataset and several datasets from UCI Machine learning repository. Effectiveness and usefulness of the proposed method is shown by comparing fitness of this method with Kmeans and Fuzzy c-means technique. For better comparative result, we have ended up with comparison of proposed clustering model with subtractive clustering (extension ...

One of the important optimization problems regarding power system issues is to determine and provide an economic condition for generation units based on the generation and transmission constraints, which is called Economic Dispatch (ED). The nonlinearity of the present problems makes conventional mathematic methods unable to propose a fast and robust solution, especially when the power system contains high number of generation units. In the present paper, an evolutionary modified Particle Swarm Optimization (PSO) is used to find fast and efficient solutions for different power systems with different generation unit numbers. The proposed algorithm is capable of solving the constraint ED problem, determining the exact output power of all the generation units. In such a way, proposed algorithm minimizes the total cost function of the generation units. To model the fuel costs of generation units, a piecewise quadratic function is used and B coefficient method is used to represent the transmission losses. The acceleration coefficients are adjusted intelligently and a novel algorithm is proposed for allocating the initial power values to the generation units. The feasibility of the proposed PSO based algorithm is demonstrated for four power system test cases consisting of 3, 6, 15, and 40 generation units. The obtained results are compared to existing results based on previous PSO implementing and Genetic Algorithm (GA). The results reveal that the proposed algorithm is capable of reaching a higher quality solution including mathematical simplicity, fast convergence, and robustness to cope with the non-linearities of economic load dispatch problem.

It is hypothesized that a key characteristic of electrocardiogram (ECG) signal is its nonlinear dynamic behaviour and that the nonlinear component changes more significantly between normal and arrhythmia conditions than the linear component. The usual statistical descriptors used in RR (R to R) interval analysis do not capture the nonlinear disposition of RR interval variability. In this paper we explore a novel approach to extract the features from nonlinear component of the RR interval signal using Teager energy operator (TEO). The key feature of Teager energy is that it models the energy of the source that generated the signal rather than the energy of the signal itself. Hence any deviations in regular rhythmic activity of the heart get reflected in the Teager energy function. The classification evaluated on MIT-BIH database, with RR interval and mean of Teager energy computed over RR interval as features, exhibits an average accuracy that exceeds 99.79%.

Digital image processing in recent decades has made considerable progress in theoretical and practical aspects. Nowadays, machine vision techniques have important application in the field of agriculture. One of these applications is detection of different varieties of rice from the bulk sample of rice image. These techniques also have high speed, accuracy and reliability. Texture feature selection is one of the important characteristics used in pattern recognition. The better feature selection of a feature set usually results in better performance in a classification problem. In This work we try to extract features by using co_occurrence matrix and select the best feature set for classification of rice varieties based on image of bulk samples using hybrid algorithm which is called "fuzzy_ imperialist competition" and then classify the best features using support vector machine(SVM). Results of the proposed method showed, the classification accuracy is improved to 96/79%. The feature set which is selected by the fuzzy-Ica provides the better classification performance compared to that obtained by Imperialist competition algorithm.

This paper proposes a method based on the Multi-Swarm Particle Swarm Optimization (PSO) with Local Search on the multi-robot search system to find a given target in a Complex environment that contains static obstacles. This method by applying Multi-Swarm with Multi-Best particles on the multi-robot system can overcome the premature convergence problem, which is one of the main problems of Basic PSO. As the time progress the global searching of the algorithm decrease and therefore the robots tend to get group together in the small-explored region that called Premature Convergence and cannot reach the target. By combining the Local Search method with Multi-Swarm, We can guarantee the global convergence of this proposed algorithm and the robots can reach the target. The Experimental results obtained in a simulated environment show that biological and sociological inspiration could be useful to meet the challenges of robotic applications that can be described as optimization problems.

An application of fuzzy modeling to the problem of telecommunications time-series prediction is proposed in this paper. The model building process is a two-stage sequential algorithm, based on Subtractive Clustering (SC) and the Orthogonal Least Squares (OLS) techniques. Particularly, the SC is first employed to partition the input space and determine the number of fuzzy rules and the premise parameters. In the sequel, an orthogonal estimator determines the input terms which should be included in the consequent part of each fuzzy rule and calculate their parameters. A comparative analysis with well-established forecasting models is conducted on real world telecommunications data, where the characteristics of the proposed forecaster are highlighted.

An application of fuzzy modeling to the problem of telecommunications time-series prediction is proposed in this paper. The model building process is a two-stage sequential algorithm, based on Subtractive Clustering (SC) and the Orthogonal Least Squares (OLS) techniques. Particularly, the SC is first employed to partition the input space and determine the number of fuzzy rules and the premise parameters. In the sequel, an orthogonal estimator determines the input terms which should be included in the consequent part of each fuzzy rule and calculate their parameters. A comparative analysis with well-established forecasting models is conducted on real world telecommunications data, where the characteristics of the proposed forecaster are highlighted.

The article demonstrates the importance of using risk analysis in the implementation of an educational-methodological project (EMP) and confirms the feasibility of using fuzzy logic for risk assessment. The use of fuzzy models allows to consider both quantitative and qualitative characteristics, as well as to represent fuzzy descriptions using fuzzy sets and linguistic variables. The fuzzy production model described (FPM) contains ten input linguistic variables, that characterize the risk factors, five output linguistic variables, that characterize the risks of EMP, including an integral risk indicator, that characterizes the quality level of the educational program in the as a whole and five rule bases. EMP allows to remove the restrictions on the number of input variables taken into consideration and to integrate both qualitative and quantitative approaches to assess the quality of methodological documents. The process of implementing the fuzzy modeling of the rule base of the specialized Fuzzy Logic Toolbox package of the MATLAB software is described. The mechanism for obtaining estimates based on the Mamdani algorithm allows getting a numerical value of the quality level, its linguistic description, as well as the degree of confidence of the expert in the resulting assessment.

Fuzzy rules regarded as a good way to represent the knowledge in many type of problems. It shorten the facts found in the problem at hand in form of IF…THEN, and the membership function regarded as basic part in the structure of these rules. The work is divided into three parts: first one is dealing with clustering process to extract the centers values; the estimation of the centers from multidimensional data set is done by using subtractive clustering algorithm. These centers are converting to fuzzy rules in the rule base, after applying a Gradient Descent method fortuning the membership function parameters. The tuning and applying Fuzzy Inference System are the second and third stage of this work. The scope of the work is heart disease diagnose.

A global image thresholding algorithm based on boundary selection is proposed for improving conventional histogram-based thresholding algorithms. An image is divided into blocks of a fixed size, and the pixel variance of each block is used for finding the boundary blocks that a bimodal histogram. Otsu's threshold value of boundary block is used to classify object blocks and background blocks. The average values of these blocks are used to determine the optimal threshold value. Experimental results show that proposed algorithm improves on conventional thresholding methods for various types of images.

An electrocardiogram (ECG) is the graphical record of bioelectric signal generated by the human body during cardiac cycle, it tells a lot about the medical status of an individual. A typical ECG waveform consist of the P, Q, R, S and T wave. The automatic ECG signal analysis comprises of using computational method/approach in extracting important features and classification of ECG waveform. This paper presents a concise ECG circuit design using an instrumentation amplifier and a band-pass passive filter. It also present the process involved in analysis of ECG signal. The first stage is the pre-filtering stage, followed by feature extraction of the signal. QRS complex is first extracted followed by P and T wave detection, also the FFT of the signal is also extracted. These features are fed into the classifier for proper classification. A pattern recognition neural network is used for classification, prior to the full deployment of the neural network, it is trained by pre-recorded ECG...

The automatic clustering is a useful tool for data-mining. It‘s a daily necessity for the searcher whatever his specialty. Indeed because of the huge amount of information available on the web-site, the access to relevant information in a suitable time is a difficult task. By grouping those informations in clusters this problem can be surmounted. Many clustering methods exist in the literature but the efficient ones suffer from some drawbacks. The main of them follows from the initialization phase which is performed randomly. Among these algorithms we find the k-means(deterministic and probabilistic version) and the clustering method based on Gaussian mixture. In these algorithms the initial parameters including the number of cluster are chosen randomly. Consequently an improper choice leads to poor clusters. In this paper we propose an approach attempting to overcome these problems. In this method the initial parameters are automatically and suitably identified. To this end, the st...

Since different varieties of crops have specific applications, it is therefore important to properly identify each cultivar, in order to avoid fake varieties being sold as genuine, i.e., fraud. Despite that properly trained human experts might accurately identify and classify crop varieties, computer vision systems are needed since conditions such as fatigue, reproducibility, and so on, can influence the expert’s judgment and assessment. Chickpea (Cicer arietinum L.) is an important legume at the world-level and has several varieties. Three chickpea varieties with a rather similar visual appearance were studied here: Adel, Arman, and Azad chickpeas. The purpose of this paper is to present a computer vision system for the automatic classification of those chickpea varieties. First, segmentation was performed using an Hue Saturation Intensity (HSI) color space threshold. Next, color and textural (from the gray level co-occurrence matrix, GLCM) properties (features) were extracted from...

There are great deals of consumer photographs which are affected by red-eye artifacts and arise frequently when shooting with flash. In this paper, a new technique is proposed to solve this problem. The proposed technique starts by detecting the skin-like regions using an optimized pixel-based neuro-fuzzy processing; morphological operations are then used to discard the extra areas after crossing the threshold. Once the skin regions are detected, five new features including geometric and color metrics are proposed to enhance the classification accuracy of the red-eye artifacts. After that, another optimized neuro-fuzzy classifier is employed to classify the red-eye regions by using the presented features. Final result is achieved by a definite syntax between skin and red-eye regions, and then, a simple correction method is used to correct the detected regions. Finally, a comparison is performed among the proposed method toward the other popular procedures and also a simple neuro-fuzzy. Final results showed the high performance of the proposed method.

Enzymes are a subclass of proteins that are specialized in catalytic activity. Protein classification problem is a difficult task because of the complexity in function and structural characteristics. This brings the necessity of computer-based methods like machine learning, artificial intelligence and data mining to solving the protein classification problems. The goal of this study is to propose the application of Fuzzy Subtractive Clustering (FSC) technique to classify the function of an enzyme by analyzing its structural class similarity to families of enzymes. A codification scheme was implemented to convert the primary structure of enzymes into a real-valued vector. To evaluate our study, the dataset obtained from Protein Data Bank (PDB) family database are used as the experimental datasets. The computational results have shown that FSC technique gives a better overall success prediction rate of 84.53% on average in comparison to previously published results.

This paper presents a tool condition monitoring approach using Takagi-Sugeno-Kang (TSK) fuzzy logic incorporating a subtractive clustering method. The experimental results show its effectiveness and satisfactory comparisons with several other artificial intelligence methods.

Enzymes are a subclass of proteins that are specialized in catalytic activity. Protein classification problem is a difficult task because of the complexity in function and structural characteristics. This brings the necessity of computer-based methods like machine learning, artificial intelligence and data mining to solving the protein classification problems. The goal of this study is to propose the application of Fuzzy Subtractive Clustering (FSC) technique to classify the function of an enzyme by analyzing its structural class similarity to families of enzymes. A codification scheme was implemented to convert the primary structure of enzymes into a real-valued vector. To evaluate our study, the dataset obtained from Protein Data Bank (PDB) family database are used as the experimental datasets. The computational results have shown that FSC technique gives a better overall success prediction rate of 84.53% on average in comparison to previously published results.