Asrul Adam

A Particle Swarm Optimization Approach to Robotic Drill Route Optimization

Most of the operational time of a PCB Robotic Drill is spent on moving the drill bit between the holes. This operational time can be kept at a minimal level by optimizing the route taken by the robot. An optimized route translates to a minimal cost of operating the robot. This paper proposes a new model that implements Particle Swarm Optimization (PSO) in order to find optimized routing path when using the PCB Robotic Drill. The main task of the PCB Robotic Drill is to drill holes at Printed Circuit Board (PCB). This PCB Robotic Drill will route the drill site by moving the drill bit along Cartesian axes from it’s initial position. Then, the drill bit will return back to the initial position. The drill route consists of a number of potential locations where the holes are going to be drilled. As the number of holes required increases so thus does the complexity to find the optimized route. The proposed model can be used to solve this complex problem with minimal computational time. The result of a case study indicates that the proposed model is capable to find the shortest path for the robot to complete its task. Thus concluded the proposed model can be implemented in any drill route problems.

A Particle Swarm Optimization Approach for Routing in VLSI

The performance of very large scale integration (VLSI) circuits is depends on the interconnected routing in the circuits. In VLSI routing, wire sizing, buffer sizing, and buffer insertion are techniques to improve power dissipation, area usage, noise, crosstalk, and time delay. Without considering buffer insertion, the shortest path in routing is assumed having the minimum delay and better performance. However, the interconnect delay can be further improved if buffers are inserted at proper locations along the routing path. Hence, this paper proposes a heuristic technique to simultaneously find the optimal routing path and buffer location for minimal interconnect delay in VLSI based on particle swarm optimization (PSO). PSO is a robust stochastic optimization technique based on the movement and information sharing of swarms. In this study, location of doglegs is employed to model the particles that represent the routing solutions in VLSI. The proposed approach has a good potential in VLSI routing and can be further extended in futureTo seek for a hyperchaotic attractor with complex topological attractor structure, a new four-dimensional continuous autonomous hyperchaotic system is proposed. Within a wider region of the variation of the control parameter, this system can generate novel hperchaotic and chaotic attractors along with quasi-periodic and periodic orbits. By employing Lyapunov exponent spectrum, bifurcation diagram, Poincaré mapping and phase portrait, etc., the existence of hyperchaotic behaviors of new system is verified and the dynamical routes from period, quasi-period, chaos and hyperchaos are observed. Furthermore, a practical circuit is designed to realize the system, which the experimental results indicate that new four-dimensional hyperchaotic system is a realizable chaotic system with potential values of engineering applications.

Feature Selection and Classifier Parameters Estimation for EEG Signals Peak Detection Using Particle Swarm Optimization

Electroencephalogram (EEG) signal peak detection is widely used in clinical applications. The peak point can be detected using several approaches, including time, frequency, time-frequency, and nonlinear domains depending on various peak features from several models. However, there is no study that provides the importance of every peak feature in contributing to a good and generalized model. In this study, feature selection and classifier parameters estimation based on particle swarm optimization (PSO) are proposed as a framework for peak detection on EEG signals in time domain analysis. Two versions of PSO are used in the study: (1) standard PSO and (2) random asynchronous particle swarm optimization (RA-PSO). The proposed framework tries to find the best combination of all the available features that offers good peak detection and a high classification rate from the results in the conducted experiments. The evaluation results indicate that the accuracy of the peak detection can be improved up to 99.90% and 98.59% for training and testing, respectively, as compared to the framework without feature selection adaptation. Additionally, the proposed framework based on RA-PSO offers a better and reliable classification rate as compared to standard PSO as it produces low variance model.

Feature selection using angle modulated simulated Kalman filter for peak classification of EEG signals

Abstract In the existing electroencephalogram (EEG) signals peak classification research, the existing models, such as Dumpala, Acir, Liu, and Dingle peak models, employ different set of features. However, all these models may not be able to offer good performance for various applications and it is found to be problem dependent. Therefore, the objective of this study is to combine all the associated features from the existing models before selecting the best combination of features. A new optimization algorithm, namely as angle modulated simulated Kalman filter (AMSKF) will be employed as feature selector. Also, the neural network random weight method is utilized in the proposed AMSKF technique as a classifier. In the conducted experiment, 11,781 samples of peak candidate are employed in this study for the validation purpose. The samples are collected from three different peak event-related EEG signals of 30 healthy subjects; (1) single eye blink, (2) double eye blink, and (3) eye movement signals. The experimental results have shown that the proposed AMSKF feature selector is able to find the best combination of features and performs at par with the existing related studies of epileptic EEG events classification.

A Hybrid Artificial Neural Network-Naive Bayes for solving imbalanced dataset problems in semiconductor manufacturing test process

This paper introduces a hybrid approach, namely Hybrid Artificial Neural Network-Naive Bayes classifier, for two-class imbalanced datasets classification. An imbalanced dataset in semiconductor manufacturing test process is chosen as a case study. Outputs prediction in semiconductor manufacturing is helpful for engineer to identify good/bad products earlier and to avoid the bad units from being processed. This application shows the significance of solving the problems. The proposed hybrid approach presented in this paper uses the concept that an Artificial Neural Network (ANN) provides a guidance to Naive Bayes classifier in making better decision by providing an additional input to Naive Bayes. Several experiments are conducted as comparison to the individual classifiers, which are ANN and Naive Bayes. As a result, the proposed Hybrid approach performs better than the individual classifiers and finally overcomes the imbalanced dataset problems in semiconductor manufacturing test process.

Feature scaling for EEG human concentration using particle swarm optimization

Electroencephalograph (EEG) is a one of recording technique that is widely used to measure human activities through brain signals. One of actively growing research in the past years is to measure human concentration using EEG. Obtaining relevant features for recognizing human concentration state becomes a challenging task due to the nature of EEG signals is a non-stationary. In the past research, various combinations of features have been employed. However, to improve the classification performance, determining the importance of each employed feature is crucially needed. In this study, feature scaling method is introduced to assign different weights for important features. Four different features are investigated in frequency domain using wavelet transform (WT). Then, particle swarm optimization (PSO) is used to scale the features while extreme learning machine (ELM) is used to classify between concentration and non-concentration states. The recorded EEG signals from Neurosky Mindwave are used to evaluate the performance of the proposed technique. The final results indicate that the proposed technique offers higher performance accuracy as compared to the methods without feature scaling.

Evaluation Of Different Peak Models Of Eye Blink Eeg For Signal Peak Detection Using Artificial Neural Network

There is a growing interest of research being conducted on detecting eye blink to assist physically impaired people for verbal communication and controlling devices using electroencephalogram (EEG) signal. One particular eye blink can be determined from use of peak points. Therefore, the purpose of peak detection algorithm is to distinguish an actual peak location from a list of peak candidates. The need of a good peak model is important in ensuring a satisfy classification performance. In general, there are various peak models available in literature, which have been tested in several peak detection algorithms. In this study, performance evaluation of the existing peak models is conducted based on Artificial Neural Network (ANN) with particle swarm optimization (PSO) as learning algorithm. This study evaluates the performance of eye blink EEG signal peak detection algorithm for four different peak models which are Dumpalas, Acirs, Lius, and Dingles peak models. To generalize the performance evaluation, two case studies of eye blink EEG signal are considered, which are single and double eye blink signals. It has been observed that the best test performance, in average, is 91.94% and 87.47% for single and double eye blink signals, respectively. These results indicate that the Acirs peak model offers high accuracy of peak detection for the two eye blink EEG signals as compared to other peak models. The result of statistical analysis also indicates that the Acirs peak model is better than Dingles and Dumpalas peak models.

Feature Selection and Classifier Parameter Estimation for EEG Signal Peak Detection Using Gravitational Search Algorithm

Peak detection is a significant step in analyzing the electroencephalography (EEG) signal because peaks may represent meaningful brain activities. Several approaches can be used for peak point detection such as time domain, frequency domain, time-frequency domain, and nonlinear approaches. The main intention of this study is to find the significant peak features in time domain approach and this can be done using feature selection methods such as gravitational search algorithm (GSA) and particle swarm optimization (PSO). This study focuses on using GSA method, a new computational intelligence algorithm. Moreover, a rule-based classifier is employed to distinguish a peak point based on the selected features. Using GSA, the parameter estimation of the classifier and the peak feature selection can be done simultaneously. Based on the experimental results, the significant peak features of the peak detection algorithm were obtained where the average test accuracy is 77.74%.

Artificial neural network — Naïve bayes fusion for solving classification problem of imbalanced dataset

Incorporating knowledge from domain expert to a classifier is one of the techniques which require to be considered in solving imbalanced dataset problems. In this study, the proposed technique is a development to extend the process for imbalanced dataset where the individual classification system has already been designed for balanced data set. This paper introduces a methodology and preliminary results which are used to investigate whether the proposed approach is possible to improve a classifiers performance when domain expert is employed to the naïve bayes classifier. Domain expert is an additional knowledge which is produced by expert system (neural network) and then become an additional input to the naïve bayes classifier. By using several benchmark data sets from the UCI Machine Learning Repository, the results of the proposed technique show an improvement as compared to the conventional naïve bayes classifier.