Siti Sophiayati Yuhaniz

Statistical Learning Theory and Support Vector Machines

It has been more than 30 years that statistical learning theory (SLT) has been introduced in the field of machine learning. Its objective is to provide a framework for studying the problem of inference that is of gaining knowledge, making predictions, making decisions or constructing models from a set of data. Support Vector Machine, a method based on SLT, then emerged and becoming a widely accepted method for solving real-world problems. This paper overviews the pattern recognition techniques and describes the state of art in SVM in the field of pattern recognition.

Support Vector Machine (SVM) for English Handwritten Character Recognition

This paper proposes a recognition model for English handwritten (lowercase, uppercase and letter) character recognition that uses Freeman chain code (FCC) as the representation technique of an image character. Chain code representation gives the boundary of a character image in which the codes represent the direction of where is the location of the next pixel. An FCC method that uses 8-neighbourhood that starts from direction labelled as 1 to 8 is used. Randomized algorithm is used to generate the FCC. After that, features vector is built. The criteria of features to input the classification is the chain code that converted to 64 features. Support vector machine (SVM) is chosen for the classification step. NIST Databases are used as the data in the experiment. Our test results show that by applying the proposed model, we reached a relatively high accuracy for the problem of English handwritten recognition.

Binary Accelerated Particle Swarm Algorithm (BAPSA) for discrete optimization problems

The majority of Combinatorial Optimization Problems (COPs) are defined in the discrete space. Hence, proposing an efficient algorithm to solve the problems has become an attractive subject in recent years. In this paper, a meta-heuristic algorithm based on Binary Particle Swarm Algorithm (BPSO) and the governing Newtonian motion laws, so-called Binary Accelerated Particle Swarm Algorithm (BAPSA) is offered for discrete search spaces. The method is presented in two global and local topologies and evaluated on the 0–1 Multidimensional Knapsack Problem (MKP) as a famous problem in the class of COPs and NP-hard problems. Besides, the results are compared with BPSO for both global and local topologies as well as Genetic Algorithm (GA). We applied three methods of Penalty Function (PF) technique, Check-and-Drop (CD) and Improved Check-and-Repair Operator (ICRO) algorithms to solve the problem of infeasible solutions in the 0–1 MKP. Experimental results show that the proposed methods have better performance than BPSO and GA especially when ICRO algorithm is applied to convert infeasible solutions to feasible ones.

Enhancement of artificial neural network learning using centripetal accelerated particle swarm optimization for medical diseases diagnosis

In recent decades, artificial neural networks (ANNs) have been extensively applied in different areas such as engineering, medicine, business, education, manufacturing and so on. Nowadays, ANNs are as a hot research in medicine especially in the fields of medical disease diagnosis. To have a high efficiency in ANN, selection of an appropriate architecture and learning algorithm is very important. ANN learning is a complex task and an efficient learning algorithm has a significant role to enhance ANN performance. In this paper, a new meta-heuristic algorithm, centripetal accelerated particle swarm optimization (CAPSO), is applied to evolve the ANN learning and accuracy. The algorithm is based on an improved scheme of particle swarm algorithm and Newton’s laws of motion. The hybrid learning of CAPSO and multi-layer perceptron (MLP) network, CAPSO-MLP, is used to classify the data of nine standard medical datasets of Hepatitis, Heart Disease, Pima Indian Diabetes, Wisconsin Prognostic Breast Cancer, Parkinson’s disease, Echocardiogram, Liver Disorders, Laryngeal 1 and Acute Inflammations. The performance of CAPSO-MLP is compared with those of PSO, gravitational search algorithm and imperialist competitive algorithm on MLP. The efficiency of methods are evaluated based on mean square error, accuracy, sensitivity, specificity, area under the receiver operating characteristics curve and statistical tests of

Embedded intelligent imaging on-board small satellites

t

An Intelligent Decision-Making System for Flood Monitoring from Space

t-test and Wilcoxon’s signed ranks test. The results indicate that CAPSO-MLP provides more effective performance than the others for medical disease diagnosis especially in term of unseen data (testing data) and datasets with high missing data values.

Detecting floods using an object based change detection approach

The strength of the adaptive neuro-fuzzy system (ANFIS) involves two contradictory requirements in a common fuzzy modeling problem, i.e. interpretability and accuracy. It is known that simultaneous optimization of accuracy and interpretability will improve performance of the system and avoid over-fitting of data. The objective of this study is the integration of particle swarm optimization (PSO) with ANFIS using modified linguistic and threshold values. This integration is expected to enhance the performance of the ANFIS system in classification problems. PSO is used to tune ANFIS parameters, to improve its classification accuracy. It is also used to find the optimal number of rules and their optimal interpretability. The proposed method has been tested on six standard data sets with different inputs of real and integer data types. The findings indicate that the proposed ANFIS–PSO integration provides a better result for classification, both in interpretability and accuracy.

Crowd analysis and its applications

Current commercial Earth Observation satellites have very restricted image processing capabilities on-board. They mostly operate according to a ‘store-and forward’ mechanism, where the images are stored on-board after being acquired from the sensors and are downlinked when contact with a ground station occurs. However, in order for disaster monitoring satellite missions to be effective, there is a need for automated and intelligent image processing onboard. In fact, the need for increasing the automation on-board is predicted as one of the main trends for future satellite missions. The main factors that hold back this concept are the limited power and computing resources on-board the spacecraft. This paper reviews existing image processing payloads of earth observing small satellites. An autonomous change detection system is proposed to demonstrate the feasibility of implementing an intelligent system on-board a small satellite. Performance results for the proposed intelligent imaging system are estimated, scaled and compared to existing hardware that are being used in the SSTL DMC satellite platform.