Junita Mohamad-Saleh

Intelligent Color Vision System for Ripeness Classification of Oil Palm Fresh Fruit Bunch

Ripeness classification of oil palm fresh fruit bunches (FFBs) during harvesting is important to ensure that they are harvested during optimum stage for maximum oil production. This paper presents the application of color vision for automated ripeness classification of oil palm FFB. Images of oil palm FFBs of type DxP Yangambi were collected and analyzed using digital image processing techniques. Then the color features were extracted from those images and used as the inputs for Artificial Neural Network (ANN) learning. The performance of the ANN for ripeness classification of oil palm FFB was investigated using two methods: training ANN with full features and training ANN with reduced features based on the Principal Component Analysis (PCA) data reduction technique. Results showed that compared with using full features in ANN, using the ANN trained with reduced features can improve the classification accuracy by 1.66% and is more effective in developing an automated ripeness classifier for oil palm FFB. The developed ripeness classifier can act as a sensor in determining the correct oil palm FFB ripeness category.

Enhanced Global-Best Artificial Bee Colony Optimization Algorithm

Artificial Bee Colony (ABC) optimization algorithm has captured much attention of researchers from various fields, in recent times. Moreover, various comparative studies clearly reports robust convergence of ABC algorithm than other bio-inspired optimization algorithms. Nevertheless, like other optimization algorithms, ABC suffers from slower convergence and tendency towards local optima trappings. Therefore, various amendments have been proposed to avertthe flaws of ABC algorithm. Nonetheless, the variants are either computationally intensive or could not avert the flaws of the algorithms. Hence, this research work proposes an efficient variant of ABC algorithm. The proposed variant capitalizes on the global-best food-source. The proposed variant has been compared with various existing variants of ABC algorithm on a few benchmark functions. Significance of the proposed variant has also been analyzed statistically. Results show the best convergence of the proposed variant among all the compared optimization algorithms on all benchmark functions.

Determination of multi-component flow process parameters based on electrical capacitance tomography data using artificial neural networks

Artificial neural networks (ANNs) have been used to investigate their capabilities at estimating key parameters for the characterization of flow processes, based on electrical capacitance-sensed tomographic (ECT) data. The estimations of the parameters are made directly, without recourse to tomographic images. The parameters of interest include component height and interface orientation of two-component flows, and component fractions of two-component and three-component flows. Separate multi-layer perceptron networks were trained with patterns consisting of pairs of simulated ECT data and the corresponding component heights, interface orientations and component fractions. The networks were then tested with patterns consisting of unlearned simulated ECT data of various flows and with real ECT data of gas–water flows. The neural systems provided estimations having mean absolute errors of less than 1% for oil and water heights and fractions and less than 10° for interface orientations. When tested with real plant ECT data, the mean absolute errors were less than 4% for water height, less than 15° for gas–water interface orientation and less than 3% for water fraction, respectively. The results demonstrate the feasibility of the application of ANNs for flow process parameter estimations based upon tomography data.

An enhanced generalized adaptive resonance theory neural network and its application to medical pattern classification

Generalized Adaptive Resonance Theory (GART) is a neural network model based on the integration of Gaussian ARTMAP and the Generalized Regression Neural Network. As demonstrated in our previous work, GART is capable of online learning and is effective in tackling both classification and regression tasks. In this paper, we propose an Enhanced GART (EGART) network whereby the capability of GART is further enhanced with the Laplacian function, a new vigilance function, a new match-tracking mechanism, and a fuzzy rule extraction procedure. The applicability of EGART to pattern classification and fuzzy rule extraction problems is evaluated using three benchmark medical data sets and one real medical diagnosis problem. The experimental results are analyzed, discussed, and compared with other reported results. The outcomes demonstrate that EGART is capable of producing high accuracy rates and of extracting useful rules for tackling medical pattern classification problems.

Color space selection for human skin detection using color-texture features and neural networks

Skin color is a robust cue in human skin detection. It has been widely used in various human-related image processing applications. Although many researches have been carried out for skin color detection, there is no consensus on which color space is the most appropriate for skin color detection because many researchers do not provide strict justification of their color space choice. In this paper, a comprehensive comparative study using the Multilayer Perceptron artificial neural network (MLP), which is a universal classifier, is carried out to evaluate the overall performance of different color-spaces for skin detection. It aims at determining the most optimal color space using color and color-texture features separately. The study has been carried out using images of different databases. The experimental results showed that the YIQ color space gives the highest separability between skin and non-skin pixels among the different color spaces tested using color features. Combining color and texture eliminates the differences between color spaces but leads to much more accurate and efficient skin detection.

An Oil Fraction Neural Sensor Developed Using Electrical capacitance Tomography Sensor Data

This paper presents novel research on the development of a generic intelligent oil fraction sensor based on Electrical capacitance Tomography (ECT) data. An artificial Neural Network (ANN) has been employed as the intelligent system to sense and estimate oil fractions from the cross-sections of two-component flows comprising oil and gas in a pipeline. Previous works only focused on estimating the oil fraction in the pipeline based on fixed ECT sensor parameters. With fixed ECT design sensors, an oil fraction neural sensor can be trained to deal with ECT data based on the particular sensor parameters, hence the neural sensor is not generic. This work focuses on development of a generic neural oil fraction sensor based on training a Multi-Layer Perceptron (MLP) ANN with various ECT sensor parameters. On average, the proposed oil fraction neural sensor has shown to be able to give a mean absolute error of 3.05% for various ECT sensor sizes.

Direct process estimation from tomographic data using artificial neural systems

The paper deals with the goal of component fraction estimation in multi-component flows, a critical measurement in many process systems. Electrical Capacitance Tomography (ECT) is an attractive sensing technique for this task, due to its low- cost, non-intrusion and fast response. However, typical systems, which include practicable real-time reconstruction algorithms have shown to give inaccurate results and the existing approaches to direct component fraction measurement have a performance that is typically flow-regime dependent, and they fail to discriminate fractions in three-component flows. Such systems also depend upon an intermediate image that must be interpreted to yield useful plant data. In the investigation described, an artificial neural network approach has been used to directly estimate the component fractions in gas-oil, gas- water and gas-oil-water flows from ECT measurements. A two-dimensional finite-element electric field model of a 12- electrode ECT sensor has been used to simulate measurements in stratified, annular and bubble-flow conditions. The singular-value decomposition has been used to reduce the raw measurement data to a mutually independent set. Multi-Layer Feed-Forward Neural Networks (MLFFNNs) have been trained with sets of such reduced ECT data with their corresponding component fractions. The trained MLFFNNs have been tested with test patterns consisting of unlearned ECT data. The paper reviews results of the best-trained networks that give a mean absolute error of less than 1% for the estimation of various multi-component fractions. The MLFFNNs’ estimations are also compared with a direct ECT method proposed in one of the previous works. The direct ECT method gives larger mean absolute errors than the MLFFNNs, demonstrating that artificial neural systems provide more accurate component fraction estimations.

Enhanced probability-selection artificial bee colony algorithm for economic load dispatch: A comprehensive analysis

The prime motive of economic load dispatch (ELD) is to optimize the production cost of electrical power generation through appropriate division of load demand among online generating units. Bio-inspired optimization algorithms have outperformed classical techniques for optimizing the production cost. Probability-selection artificial bee colony (PS-ABC) algorithm is a recently proposed variant of ABC optimization algorithm. PS-ABC generates optimal solutions using three different mutation equations simultaneously. The results show improved performance of PS-ABC over the ABC algorithm. Nevertheless, all the mutation equations of PS-ABC are excessively self-reinforced and, hence, PS-ABC is prone to premature convergence. Therefore, this research work has replaced the mutation equations and has improved the scout-bee stage of PS-ABC for enhancing the algorithms performance. The proposed algorithm has been compared with many ABC variants and numerous other optimization algorithms on benchmark functions and ELD test cases. The adapted ELD test cases comprise of transmission losses, multiple-fuel effect, valve-point effect and toxic gases emission constraints. The results reveal that the proposed algorithm has the best capability to yield the optimal solution for the problem among the compared algorithms.

A Q-learning-based multi-agent system for data classification

Display Omitted A multi-agent classifier system with Q-learning for data classification is proposed.The Trust-Negotiation-Communication reasoning scheme for agent teaming is utilized.A trust measurement that combines Q-learning and Bayesian formalism is formulated.Benchmark problems with and without noise are used to evaluate the proposed model.The results show that the proposed model is useful for data classification problems. In this paper, a multi-agent classifier system with Q-learning is proposed for tackling data classification problems. A trust measurement using a combination of Q-learning and Bayesian formalism is formulated. Specifically, a number of learning agents comprising hybrid neural networks with Q-learning, which we have formulated in our previous work, are devised to form the proposed Q-learning Multi-Agent Classifier System (QMACS). The time complexity of QMACS is analyzed using the big O-notation method. In addition, a number of benchmark problems are employed to evaluate the effectiveness of QMACS, which include small and large data sets with and without noise. To analyze the QMACS performance statistically, the bootstrap method with 95% confidence interval is used. The results from QMACS are compared with those from its constituents and other models reported in the literature. The outcome indicates the effectiveness of QMACS in combining the predictions from its learning agents to improve the overall classification performance.

New Enhanced Artificial Bee Colony (JA-ABC5) Algorithm with Application for Reactive Power Optimization

The standard artificial bee colony (ABC) algorithm involves exploration and exploitation processes which need to be balanced for enhanced performance. This paper proposes a new modified ABC algorithm named JA-ABC5 to enhance convergence speed and improve the ability to reach the global optimum by balancing exploration and exploitation processes. New stages have been proposed at the earlier stages of the algorithm to increase the exploitation process. Besides that, modified mutation equations have also been introduced in the employed and onlooker-bees phases to balance the two processes. The performance of JA-ABC5 has been analyzed on 27 commonly used benchmark functions and tested to optimize the reactive power optimization problem. The performance results have clearly shown that the newly proposed algorithm has outperformed other compared algorithms in terms of convergence speed and global optimum achievement.