22 research outputs found
Musical instrument identification using Convolutional Neural Network (CNN) algorithm / Muhammad Nur Azri Irfan Abdul Rahman
No full textThe motivation behind the project was to help automate the cumbersome task of validating instruments from images using Convolutional Neural Network (CNNs) algorithm to identify the musical instrument so that this task could be completed with higher accuracy. This approach tried to overcome the limitations of the manual method and traditional algorithm, which tends to fail with the diverse dataset, diverse visual features, and scalability. The methodology followed a structured three-phase process: The first stage was the collection of a dataset of 5,099 images of 30 different musical instruments of Kaggle, providing variable lighting, angles, or backgrounds, along with preprocessing to standardize the inputs. In the development phase, Convolutional Neural Network model was designed and trained using sophisticated techniques of data augmentation, dropping out and hyperparameter tuning under the supervised learning methodology to increase the performance of the system. Finally, the rigor of evaluation phase is carried out to evaluate the model utilizing precision, recall, F1 score, and the overall accuracy metrics which ascertained robustness and reliability for the model
Kerangka konstruktif perbandingan untuk jurugambar kekurangan penglihatan warna dalam fotografi digital / Muhammad Azri Abdul Rahman and Nadzri Mohd Sharif
No full textKekurangan Penglihatan Warna (KPW) adalah keadaan genetik yang paling lazim, mempengaruhi satu daripada 20 lelaki dan satu daripada 200 perempuan. Jurugambar fotografi digital menerima impak yang besar akibat daripada kekurangan ini. Individu yang menghidap KPW tidak dapat menghayati karya seni kerana mereka mengalami kekeliruan terhadap warnanya. Hasilnya, penyelidik menjalankan kajian dalam mencipta kerangka kerja yang berkesan untuk lebih mengklasifikasikan keadaan penglihatan jurugambar KPW dengan menggunakan kaedah berasaskan RSk (Rona, Saturasi, Kecerahan), pelarasan kontor, dan proses interpretasi ke dalam fotografi. Kesan perbezaan karya fotografer normal dan KPW dan dilihat dari aspek keaslian warna, ketekalan warna dan kontras warna. Teknik kaedah yang diimplementasikan di dalam kajian ini adalah kaedah eksperimen di mana persampelan dipilih bertujuan oleh pengkaji. Sampel mematuhi syarat yang ditetapkan dan mempunyai teknik dan kemahiran yang sama. Terdapat 40 orang responden (n=40), dimana sampel kemudiannya dibahagikan kepada dua kumpulan iaitu kumpulan kawalan dan rawatan dengan 20 orang responden (n=20) untuk setiap kumpulan, dan mereka dinilai dengan soalan dan imej yang sama. Sampel kawalan diuji menggunakan gambar normal dan mempunyai arahan yang lengkap, manakala sampel rawatan diuji dengan gambar KPW dan kekurangan arahan. Antara soalan yang terdapat dalam soalan adalah kaedah berasaskan RSk (Rona, Saturasi, Kecerahan), peningkatan kontor dan proses tafsiran. Penemuan penyelidikan mampu membantu jurugambar KPW untuk mencapai persepsi warna yang lebih tepat
Overconfidence bias among investors: A qualitative evidence from Ponzi scheme case study
No full textThis study aims to examine the prevalence of overconfidence bias in the decision-making process of Malaysian investors in Ponzi schemes. We explore a well-documented behavior that distorts the investor’s judgment, leading to a future event’s miscalculation — a psychological bias known as overconfidence bias (Kuranchie-Pong & Forson, 2022). Our study offers a novel viewpoint by investigating the hard-to-reach type of investor, the Ponzi scheme investors using the behavioral finance theory and qualitative method. Therefore, this investigation employed qualitative reasoning, which could also be an example of applying thematic analysis using ATLAS.ti. This study’s findings indicate that Ponzi scheme investors exhibit overconfidence bias in investing in the Ponzi investment schemes. We unraveled three types of overconfidence bias that prevail in the Ponzi scheme investors’ decision process. Acknowledging its limitations as a qualitative inquiry, the authors call for a joint effort to explore this field of study further. This emerging area of investor behavior research will afford valuable knowledge that could resolve the mysteries behind the never-ending issue of the Ponzi investment scheme
First report of pathogenic <i>Leptospira</i> spp. isolated from urine and kidneys of naturally infected cats
No full textLeptospirosis is one of the most widespread zoonotic diseases and can infect both humans and animals worldwide. Healthy cat, as a potential source of exposure to humans, are likely underestimated owing to the lack of overt clinical signs associated with Leptospira spp. infection in this species. The aim of the study was to determine the exposure, shedding, and carrier status of leptospires in shelter cats in Malaysia by using serological, molecular, and bacteriological methods. For this study, 82 healthy cats from two shelters were sampled. The blood, urine, and kidneys were tested using the microscopic agglutination test (MAT), polymerase chain reaction (PCR), and bacterial culture. On the basis of serological, molecular, and/or culture techniques, the total detection of leptospiral infection was 29.3% (n = 24/82). Through culture techniques, 16.7% (n = 4/24) of the cats that tested positive were carriers with positive kidney cultures, and one cat was culture positive for both urine and kidney. The Leptospira spp. isolates were identified as pathogenic L. interrogans serovar Bataviae through serological and molecular methods. Through serological techniques, 87.5% (n = 21/24) had positive antibody titers (100–1600) and most of the Bataviae serogroup (n = 19/21). Using PCR, 16.7% (n = 4/24) of cats were shown to have pathogenic Leptospira spp. DNA in their urine. Furthermore, three out of four culture positive cats were serology negative. The present study reports the first retrieval of pathogenic leptospires from urine and kidneys obtained from naturally infected cats. The results provide evidence of the potential role of naturally infected cats in the transmission of leptospires. Additionally, leptospiral infection occurs sub-clinically in cats. The culture isolation provides evidence that healthy cats could be reservoirs of leptospiral infection, and this information may promote the development of disease prevention strategies for the cat population.</div
Silicon Nanowires Based DNA Biosensor
No full textProf. Dr. Uda Hashim and his team won bronze at Malaysia Technology Expo 2010 (MTE 2010), 4th - 6th February 2010 at Putra World Trade Centre (PWTC), Kuala Lumpur. MTE 2010 organized by MARS and PROTEMP Exhibitions Sdn. Bhd.This research describes the fabrication of silicon nanowire for DNA hybridization detection in biosensor application. The SiNWs are fabricated by electron beam lithography and followed by deposition of contact metals. Aurum is a suitable metal to make Ohmic contact with SiNWs and its exhibit electrical transport measurements of the devices. The presence of the hybridization DNA will induce electronic charge and changed the electronic properties of the devices from which detectable electronic signals are generated
3D modelling for multipurpose cadastre
No full textThree-dimensional (3D) modelling of cadastral objects (such as legal spaces around buildings, around utility networks and other spaces) is one of the important aspects for a multipurpose cadastre (MPC). This paper describes the 3D modelling of the objects for MPC and its usage to the knowledge of 3D cadastre since more and more related agencies attempt to develop or embed 3D components into the MPC. We also intend to describe the initiative by Malaysian national mapping and cadastral agency (NMCA) in developing the 3D MPC. In this case, we strongly believe that 3D city modelling approach could be utilized for such purposes, e.g. linking the physical and legal spaces of buildings and other objects within the multipurpose cadastre as well as for developing the relevant database, thus, provide queries to the objects. Other relevant information and services also could be developed for other layers within MPC for different applications or services. Basically, the Malaysian MPC consists of large scale spatial data layers grouped into two types of sub systems. The first group consists of components with spatial objects to which legal and administrative facts are attached (rights, restrictions and responsibilities), such as 3D cadastre (volume parcels), 3D marine cadastre, and 3D strata (in buildings). The second group consists of components with reference to physical spatial objects, such as: 3D topography (with buildings footprints), underground utility (3D), and 3D city model (with roof structure and LoD3 buildings) where each of these components will be elaborated and discussed in this paper. As one of the components for the proposed Malaysian MPC, 3D models of cadastral objects are needed to support the existing cadastre system by providing spatial and semantic information of the 3D objects according to the exact situation as in the real world. The representation of the real world objects in 3D form will enable the concerned parties to understand the situation and problems related to the particular land site as well as increasing the efficiency in managing land-related matters. The concept for 3D volume parcels for 3D cadaster and 3D modelling with underground utility networks is presented. This paper relates the Malaysian MPC model with the LADM (Land Administration Domain Model) where the generation of the UML model that complies with the concept of LADM for the Malaysian MPC is addressed. We anticipate that the proposed approach for the MPC could be implemented by the authority in the near future. Outlook for future works and the emerging 3D modelling technologies for the MPC then form a part of conclusions of this paper.OTB ResearchOTB Research Institute for the Built Environmen
Validation of a scale for the measurement of employee competency in relation to succession planning amongst administrators in higher education institutions
No full textPurpose: This paper examines the validity and reliability of an instrument to evaluate employee competency in the context of succession planning using a sample of administrators within the 48 to 54 grades in Malaysian public universities. Theoretical framework: Shaping the qualities and competencies of university administrators can serve as a model of competent leadership for the sustainability of succession planning for higher education institutions. Design/methodology/approach: Systematic random sampling was used for data collection using questionnaires, including pilot studies and subsequent field studies. The survey was administered to 430 respondents from the administrators of 20 public universities. The study conducted an exploratory factor analysis (EFA) using the pilot study data (n = 102) to identify potential factor structures. Confirmatory factor analysis (CFA) using field study data (n = 328) was conducted to confirm the construct measurement model. Findings: The EFA and CFA validation processes suggested that only 10 modified items with 4 components remained to achieve a better fit model: Interpersonal (IP), Personal Effectiveness (PE), Teamwork (TW), and Self-Development (SD). All composite reliability (CR) measures for these 4 components exceeded the required value of 0.6. Since the values of these indices fall within the range of their respective cutoff values, the model meets the goodness-of-fit criterion: RMSEA = 0.070,GFI=0.955, CFI = 0.953, TLI = 0.926, and a normalized chi-square = 2.605. According to the study, the instrument is valid and reliable for assessing the components of the employee competency model. Research, Practical & Social implications: The scale validated in this study could be used by human resource managers in public universities as a tool to better understand the belief systems of their employees. It could help them make an informed decision that will enable them to provide competent administrators for succession planning. Originality/value: This study is a catalyst to assess and highlight the current status and impact of employee competency on succession planning in Malaysian higher education institutions
A Constructive Framework for Colour Vision Deficiency (CVD) in Digital Photography
No full textStudies show that around 8% of men and about 0.5% of women globally are affected by color blindness (Nathans et al., 1993; Neitz & Neitz, 2011), making the appreciation of photography challenging. Further research is required to explore the visual difficulties that color vision deficiency (CVD) photographers face. This study evaluates the visual skills of CVD photographers, explicitly focusing on the categories of Protanopia, Tritanopia, and Deuteranopia. It employs the HSx-Based Method, Contour Adjustment, Interpretation Process, and Perception Learning. After preliminary testing, experimental testing was conducted with control questions and minimal coaching. The Chi-square Test of Homogeneity classifies CVD and compares it to normal vision. Spearman's Rho Correlation Test confirms that all CVD categories are "photosafe," providing valuable software and product development insights
Serological detection of anti-leptospira antibodies among animal caretakers, dogs and cats housed in animal shelters in Peninsula Malaysia
No full textLeptospirosis is a worldwide zoonotic disease caused by Leptospira spp. and it has emerged as one of the major public health issues worldwide. Despite extensive studies, information regarding leptospirosis in animal shelters in Malaysia is still scarce. The aim of this study was to determine anti-Leptospira antibodies and the most common circulating serogroups among humans and animals in animal shelters in west and south parts of Peninsular Malaysia. Blood samples were obtained from 58 humans, 127 dogs, and 47 cats, that were recruited from two shelters. All humans and dogs appeared healthy, except few cats showed clinical signs of mild feline upper respiratory disease. Microscopic Agglutination Test (MAT) was used to detect anti-Leptospira antibodies against 20 pathogenic serovars. Based on the cut-off antibody titre ≥ 1:100, the sero-detection of Leptospira spp. in human, dogs, and cats were as 8.62%, 20.47%, and 14.89%, respectively. Serogroup Bataviae was found in human, dogs and cats, whereas Bataviae, Javanica, and Ballum were the common serogroups among dogs and cats. Anti-Leptospira antibodies titres were in the range from 1:100 to 1:200 in human and 1:100 to 1:400 in sheltered animals. Sero-detection studies of anti-Leptospira antibodies in shelters environment worldwide and regionally is necessary to increase the public health awareness and to understand the risk of this zoonotic disease. Furthermore, data regarding the predominant serogroups is needed in a local setting for further vaccination development studies. More studies are warranted to investigate the role of sheltered animals in leptospiral transmission in its environment
