Maleerat Sodanil

Prediction of stock price using an adaptive Neuro-Fuzzy Inference System trained by Firefly Algorithm

The substance of the design of Adaptive Neuro-Fuzzy Inference System (ANFIS) can be seen as an optimization problem to find the best parameters with minimal error function. This paper proposes a combination of the Firefly Algorithm and Adaptive Neuro-Fuzzy Inference System. The fuzzy neural network model will be trained by the Firefly Algorithm, and applied to predict stock prices in the Vietnam Stock Market. The experiments will compare performance between the proposed system and ANFIS trained by the Hybrid Algorithm, Back Propagation and Particle Swarm Optimization (PSO). The experimental results show that the system has reasonable efficient performance.

A Knowledge Transfer Framework for Supporting the Transition to Agile Development of Web Application in the Thai Telecommunications Industry

Agile software development methods are often applied in volatile software development environments typically perceived as being difficult to tackle by traditional methods. Yet, only a minority of organizations is able to transfer to agile development effectively. This paper proposes a knowledge transfer framework supporting the transition to agile development with guidance on how to put knowledge transfer into action. In this framework, a knowledge transfer process consists of six components (i.e., problems, antecedents, knowledge, mechanisms, knowledge application, and outcomes) and flows through four distinct stages (i.e., Initiation, Implementation, Ramp-up, and Integration). In each stage, components interact with each other multi-directionally and play an important role depending on the stages functionality. A set of knowledge transfer activities in each stage is also specified which aligns with agile ways, especially Scrum. The description of the application of the developed framework and lessons learned conclude the paper.

Artificial neural network-based time series analysis forecasting for the amount of solid waste in Bangkok

Solid waste is a municipal environmental problem which difficult to manage. Thus, a solid waste forecasting model is essential for the effective management and planning. This paper aims to develop a time series forecasting model for the amount of solid waste generated in Bangkok using artificial neural networks, and offers a suitable model for solid waste forecasting. The time series data were collected as monthly accounts of solid waste generated between October 2002 and July 2013. Then, the data were cleaned and converted in order to accurately analyze. The forecast model was developed using predictive analytic tool Rapidminer. Artificial neural network model was trained with backpropagation algorithm. The results showed that the network structure of 3-35-1 performs the greatest performance with prediction accuracy at 0.870 and MSE equaling 0.2333.

An improved local binary pattern for edge detection of images

This research proposed the improved H-LBP for edge detection, TH-LBP. It applied the Hyperbolic Tanh function instead of the H function. In this method, the continuous and thick edges can be obtained. In addition, the method parameters could be adjusted to increase the edge details in the low contrast area. This method can solve the problem in the case that the two different local areas obtain the same result.

Application of logarithm, wavelet and contourlet transform for video-frame quality improvment

Video shot boundary detection is important step for the research in the content analysis and retrieval fields. In this paper, firstly we presented efficient method for Video-frame quality improvement to suppress flash occurred within video frame using logarithm, wavelet and contourlet transform. In addition, wavelet and contourlet transform also performed denoising. Secondly, for shot boundary detection, we used gray-scale histogram differences with edge change ratio. Furthermore, the adaptive threshold algorithm for shot transition detection was proposed. The experiment results showed that using logarithm with contourlet transform gain the precision and recall higher than using logarithm with wavelet transform.

Ontology knowledge-based framework for machine learning concept

In the objective of this paper was to present ontology knowledge-based design and development to explain concepts and machine learning techniques which were compiled from book, articles, research and websites that publish information. The database structure includes 4 application domains: 1) learning 2) learning techniques 3) learning evaluation and 4) machine learning technique applications. The experimental evaluation was conducted by retrieving data using question sets. The results of the evaluation showed precision value at 99.65 percent and recall value at 95.90 percent. This machine learning ontology could be applied to other related information systems and databases for future development and further research.

A knowledge transfer framework for secure coding practices

Building a secure software product is required understandings of security principles and guidelines for the secure coding in terms of programming languages to develop safe, reliable, and secure systems in software development process. Therefore, knowledge transferring is required and influenced to the most effective secure software development project. This paper proposes a knowledge transfer framework for secure coding practices with guidance for the development of secure software product and how the framework could be applied in the telecommunication industry. A set of knowledge transfer activities is specified which aligns for secure coding. Finally, the implementation of a knowledge transfer framework for secure coding practices could mitigate at least the most common mistakes in software development processes.

A Development of Image Enhancement for CCTV Images

This research aims to propose the method of the CCTV image enhancement using sub-images homomorphic filtering techniques. The experiments were conducted using CCTV video clips which recorded from both indoor and outdoor surveillance video system. The processes started by dividing video clip into image frames, then separate both horizontal and vertical parts before enhancing with homomorphic filtering methods. The results given quite good performance in terms of PSNR. Therefore the proposed method based on homomorphic filtering techniques using sub-images dividing can be used to improve the quality of CCTV images.

An Ontology-Based Query Expansion for an Agricultural Expert Retrieval System

Query expansion is a technique of information retrieval system which considered to the context of the users queries in order to improve the retrieval effectiveness. There are several method expansion methods have been investigated. An ontology-based approach is one of the expansion approaches which used as knowledge-based for searching. This paper proposed an ontology-based query expansion using a concept of the combination of an IR technique and association rule mining to examine and evaluate of an agricultural expert retrieval systems. An association rule mining is applied in the inference engine in order to optimize the users queries. A set of inference rules are also created to support the expertise retrieval task. The result set is depends on the new query which expands from users queries as keywords using an agricultural ontology structures. The experts who have expertise in topics or keywords, type of plants and problem solving are the elements of the ontology and will be used to search the relevant publications. The experiments were conducted using publications from collections of the Thai National AGRIS center. The results show that the improvement of this expansion method yields better performance than using basic query expansion search with the F-measure equal to 98.87%.