Papers by Safie mat yatim
A Review on Real-Time Terrain Rendering Techniques
Abstract: Practical pattern classification problems require selection of a subset of attributes o... more Abstract: Practical pattern classification problems require selection of a subset of attributes or features to represent the patterns to be classified. The feature selection process is very important which selects the informative features for used classification process. This is due to the fact that performance of the classifier is sensitive to the choice of the features used to construct the good classifier from small or high dimension data that are inherently noisy. In this paper, we propose an efficient feature selection method that finding and selecting ...
Pembangunan pangkalan data demografi dalam konteks sistem maklumat grafik (GIS)
Demographic information is an important source to the administrators and planners in determining ... more Demographic information is an important source to the administrators and planners in determining the living status of the population in a particular area. This information act as a guideline in determining the planning priority that need to be implemented for that area. However the main problem frequently faced by the administrators and planners in most developing countries like Malaysia is the difficulty in getting a complete and up-to-date information. The most effective effort to overcome the problem is by developing a database system which enable a complete and up-to-date demographic information based on the division of area (for example, administrative border) being kept and accessed.
Accuracy in rendering sky/atmosphere color has long been neglected and reduced to fictitiously si... more Accuracy in rendering sky/atmosphere color has long been neglected and reduced to fictitiously simplified methods such as uniform blue color, color gradation from zenith to horizon and static textured sky dome. These methods may save up a lot of CPU-computation time; nonetheless they lack the ability of working with arbitrarily complex atmospheric conditions. As the light itself is a fraction of the electromagnetic spectrum, we need a spectral-based solution to compute the resulting color of the sky. This is being the result of interaction processes between light and the atmosphere, namely absorption and scattering being wavelength-dependent. We present a fast yet accurate model for approximating full spectrum daylight for local atmospheric conditions based largely on radiative transfer works in atmospheric science. Our aim is to design an inexpensive model yet maintain the physical-based radiometry to a certain level without sacrificing accuracy and usability for outdoor computer g...
Permonitoran dan kawalan projek : pendekatan analisis rangkaian
Planning, scheduling, monitoring, and control are considered to be the utmost important phases in... more Planning, scheduling, monitoring, and control are considered to be the utmost important phases in project management Project planning and scheduling methodology developed based on network analysis are proven to produce the better outcome in acheiving a more efficient project handling. Despite this, the development of the methods for control and generation monitoring information from network analysis were not given the appropriate attention compared to the development of planning and scheduling methods. This paper discussed the facilities that can be offered by network analysis. Discussion will be focused on the building of control mechanism or system, control methods, and the monitoring information that can be generated by network analysis in order to produce a more efficient control on project handling and implementation.
Practical pattern classification problems require selection of a subset of attributes or features... more Practical pattern classification problems require selection of a subset of attributes or features to represent the patterns to be classified. The feature selection process is very important which selects the informative features for used classification process. This is due to the fact that performance of the classifier is sensitive to the choice of the features used to construct the good classifier from small or high dimension data that are inherently noisy. In this paper, we propose an efficient feature selection method that finding and selecting informative features from small or high dimension data which maximum the classification accuracy. In this work, we apply genetic algorithm to search out and identify the potential informative features combinations for classification and then use the classification accuracy from the support vector machine classifier to determine the fitness in genetic algorithm. Experimental results with benchmark datasets show the usefulness of the propose...
Contour Lines Recognition from Raster Topographic Maps
Map digitization is still an ongoing research in development of GIS, photogrammetry, remote sensi... more Map digitization is still an ongoing research in development of GIS, photogrammetry, remote sensing, military, urban planning and etc. Majority of recent approach, either associates with human interaction or semi-automated process are insufficient in various aspects. There are usually time-consuming, inefficient, and shown inaccuracies in digitized data. This research is intended to develop an efficient approach for contour lines recognition from scanned topographic maps. A detail investigation of existing automated map digitization process based on formal, heuristic, knowledge and interactive algorithm is carried out. Several vectorization and image acquisition techniques are performed at early stage to analyze the map for ensuring all features on map being well separated, located, classified, and traced, consequently stored in vector format. The overlaid information and texture background are distinguished and segmented by color quantization approach. Due to scanning or deformatio...
Hubungan masa dan kos projek : algoritma dan kategori kos
This paper discusses network analysis approach in analysing activities and projects costs. The re... more This paper discusses network analysis approach in analysing activities and projects costs. The relationship between activities time and cost, and the algorithm which determine scheduling times for each activity with the least project cost. Furthermore, the relationship between costs, resources and time of the project will also be discussed.
The development of machine learning based software for predicting protein-protein interactions and protein function from protein primary structure
Understanding proteins functions is a major goal in the post-genomic era. Proteins usually work i... more Understanding proteins functions is a major goal in the post-genomic era. Proteins usually work in context of other proteins and rarely function alone. Therefore, it is highly relevant to study the interaction partners of a protein in order to understand its function. For this reason, the main objective of this thesis is to predict protein-protein interactions based only on protein primary structure. Using the Support Vector Machines (SVM), different protein features have been studied and examined. These features include protein domain structures, hydrophobicity and amino acid compositions. The results imply that the protein domain structure is the most informative feature for predicting protein-protein interactions. It also requires much lower running time compared to the other features. However, using normal binary SVM requires positive and negative data samples. Although it is easy to get a dataset of interacting proteins as positive examples, there are no experimentally confirme...
The development of multi-agent system for biological information retrieval from heterogeneous distributed data and knowledge sources
Distribution, complexity, heterogeneity of resources, software, and hardware are challenges in di... more Distribution, complexity, heterogeneity of resources, software, and hardware are challenges in distributed applications. Intelligent mobile agent is used to handles distributed data, control, expertise, and resources in distributed applications. It is an executing program that is capable of flexible autonomous action, migrates during execution from host to host, reason with uncertain, incomplete, and contradictory information. Belief-Desire-Intention-Joint-Mobile UNITY (BDI-MU) kit is introduced to construct intelligent mobile agent. It is designed using BDI-MU architecture. BDI-MU architecture is an open architecture based on autonomous practical reasoning and social ability through messages. It consists of system package and user-defined package. The system package provides components that can be extended and imported to develop an agent, its server and domain. The user-defined package is a pluggable component which consists of user applications, agents and databases. The BDI-MU a...
Spesifikasi bahasa pengaturcaraan AMAL
Programming languages have become an important tool for communication between human and computers... more Programming languages have become an important tool for communication between human and computers. The generation of computer languages has evolved, starting from the first generation in the form of machine language to the sixth generation, which is prone to natural language. This paper will highlight a specification of a new programming language, called AMAL. The language constructs of this programming language use Bahasa Malaysia and is based on the existing programming languages Pascal, Ada and C.
View Frustum Culling for Three-Dimensional Virtual Geographic Information System
Visibility culling is a crucial task for managing complex structure of three-dimensional scene es... more Visibility culling is a crucial task for managing complex structure of three-dimensional scene especially in Geographic Information System (GIS) domain. According to the recent hardware specification, it is impractical to render large amount of dataset in real-time environment due to the quantity of dataset need to be faced exceeds the capability and performance of computer’s physical memory. Therefore, view frustum culling has been proposed in order to eradic ate the non-visible objects from pyramid of vision and subsequently speed up the system during the runtime process. It is based on predictive hierarchical quadtree structure and intersection test is done by calculating the distance between bounding volume of the object’s surfaces and the view volume. In this paper, we will adopt and adapt this approach to the 3D terrain objects for our deve lopment of 3D Virtual GIS.
Research Issues in Development of Three-Dimensional Virtual Geographical Information System
The evolvement of hardware technologies leads the growth in development of Geographical Informati... more The evolvement of hardware technologies leads the growth in development of Geographical Information System (GIS) applications that represent the geographic space in a digital environment. The conventional GIS are unable to handle the complex nature of the spatial data. Migration and transition from 2D to 3D emerge some problems in order to construct robust GIS applications. Two core obstacles, that is, database management and fast real-time visualization are main difficulties that encounter nowadays. However, the performance of data operation on the huge multi-dimensional database is still a difficult subject, especially involve in real-time visualization. Users and developers perspectives have to be considered and fulfilled in order to remedy the problems. A number of researchers have proposed their own optimization algorithms, techniques and approaches which attempt to reduce the complexity of the problems. A survey has been made to review the current issues related to the development of 3D Virtual GIS.
Three-Dimensional Terrain Database Design and Management for Development of Virtual Geographical Information System
Real-time terrain rendering and visualization is an interesting research area due to its capabili... more Real-time terrain rendering and visualization is an interesting research area due to its capabilities to provide a challenging environment for a number of researchers and interactive application’s developers. In threedimensional geographical information system (3D GIS) domain, they need to maintain interactive frame rate continuously as well as preserve the visual fidelity and cope large amount of various databases. In this paper, a set of methods has been designed and proposed for handling massive terrain databases in order to minimize the computation time for real-time processing and avoid rendering redundant geometry. Hence, terrain databases need to be collected and then converted into single internal representation as our own hierarchical quadtree data structure. To allow arbitrary resolution of terrain database to be loaded and rendered, an algorithm for verifying the dynamic size of database will be adopted in pre-processing level. We designed two core components in managing terrain databases which are view frustum culling (block-based) to discard the unseen terrain objects that are not in view volume; and view-dependant level of detail (vertex-based) to give polygon details on appropriate terrain area that need it. A system will be developed depending on all these methods and the measurement will be concerned on visualization accuracy, frame rate (frame per second) and polygon generated for each frame.
Pengindeksan dan capaian dokumen dalam pangkalan data teks bebas : satu pendekatan
Free text database is considered something unusual and different from conventional databases. Rec... more Free text database is considered something unusual and different from conventional databases. Records in this database are documents which have no fixed structure or forms. Hence, databases of this kind requires a different kind of data or documents manipulation because of the unconventionality of the techniques used in indexing the database. This paper discusses all process which are involved in building a large free text database management system. In addition of that, documents indexing and retrieval techniques will also be discussed.
Hubungan masa dan kos projek : algoritma dan kategori kos
A study on biomolecular sequence alignment using machine learning techniques
Pembangunan pangkalan data demografi dalam konteks sistem maklumat grafik (GIS)
The development of machine learning based software for predicting protein-protein interactions and protein function from protein primary structure
Page 1. VOT 74288 THE DEVELOPMENT OF MACHINE LEARNING BASED SOFTWARE FOR PREDICTING PROTEIN-PROTE... more Page 1. VOT 74288 THE DEVELOPMENT OF MACHINE LEARNING BASED SOFTWARE FOR PREDICTING PROTEIN-PROTEIN INTERACTIONS AND PROTEIN FUNCTION FROM PROTEIN PRIMARY STRUCTURE (PEMBANGUNAN ...
The Development Of Integrated Planning And Scheduling Framework For Dynamic And Reactive Environment Of Complex Manufacturing Problem
... 51 4.5 Particle Swarm Optimization 53 Page 6. vi ... 1.9 Suggestion for Further Work We propo... more ... 51 4.5 Particle Swarm Optimization 53 Page 6. vi ... 1.9 Suggestion for Further Work We propose to use GA with a match-up approach to solve dynamic problem in the job shop scheduling problem. GA was chosen since it is well suited to optimization problem and ...
Uploads
Papers by Safie mat yatim