Search

Training an imbalanced dataset can cause classifiers to overfit the majority class and increase the possibility of information loss for the minority class. Moreover, accuracy may not give a clear picture of the classifier’s performance. This paper utilized decision tree (DT), support vector machine (SVM), artificial neural networks (ANN), K-nearest neighbors (KNN) and Naïve Bayes (NB) besides ensemble models like random forest (RF) and gradient boosting (GB), which use bagging and boosting methods, three sampling approaches and seven performance metrics to investigate the effect of class imbalance on water quality data. Based on the results, the best model was gradient boosting without resampling for almost all metrics except balanced accuracy, sensitivity and area under the curve (AUC), followed by random forest model without resampling in term of specificity, precision and AUC. However, in term of balanced accuracy and sensitivity, the highest performance was achieved by random forest with a random under-sampling dataset. Focusing on each performance metric separately, the results showed that for specificity and precision, it is better not to preprocess all the ensemble classifiers. Nevertheless, the results for balanced accuracy and sensitivity showed improvement for both ensemble classifiers when using all the resampled dataset.

Dengue fever is a well-known vector-borne disease caused by Aedes aegypti mosquito. It has become a major burden to economy and society of affected country. In Malaysia, dengue incidence in Selangor has been worsening and alarming. The aim of this study is two-fold; to examine the trend and behaviour of dengue incidence across time and districts in Selangor and to cluster the endemic areas in Selangor using Wards hierarchical clustering method. The spatial and temporal analysis found that the dengue incidence is worsening in the early and middle of the year. The Wards minimum variance method was able to cluster Selangor's endemic area into high endemic areas (Gombak, Hulu Langat, Klang and Petaling), medium endemic area (Sepang) and low endemic areas (Hulu Selangor, Kuala Langat, Kuala Selangor, SabakBernam). The findings of the study are significant to respective local authorities in providing information for monitoring and planning the early dengue warning systems. This is important to reduce the dengue incidence in hot spot areas and to safeguard the community from dengue outbreak.

Corporate social responsibility (CSR) has become an imperative tool to address challenges and achieve sustainable growth. Realizing its impact to the society, companies are demanded to participate in sustainable development of which poverty eradication is one of it. The CSR practice, to date, is not strategically planned and executed especially when it comes into philanthropic corporate social responsibility (PCSR). This could be due to failure to identify categories of PCSR activities, limiting its effectiveness to achieve the intended outcomes. Thus, document clustering is proposed to be used to automate the pattern identification process. This study has extended document clustering by integrating the traditional document clustering application with topic modeling approach. This integrated approach enables the identification of the PCSR pattern. The analysis involved a three-year data from the annual report of the 25 CSR-award winning companies in Malaysia which involved several steps. Findings from this study revealed seven clusters that represented seven types of PCSR activities performed by the CSR-award winning companies in Malaysia. The findings offer an insight to be considered by companies in strategizing the CSR activities, particularly philanthropic responsibility in ensuring optimum impact to innovatively support the society and contribute towards poverty mitigation.

Surveillance is a critical component of any dengue prevention and control program. There is an increasing effort to use drones in mosquito control surveillance. Due to the novelty of drones, data are scarce on the impact and acceptance of their use in the communities to collect health-related data. The use of drones raises concerns about the protection of human privacy. Here, we show how willingness to be trained and acceptance of drone use in tech-savvy communities can help further discussions in mosquito surveillance. A cross-sectional study was conducted in Malaysia, Mexico, and Turkey to assess knowledge of diseases caused by Aedes mosquitoes, perceptions about drone use for data collection, and acceptance of drones for Aedes mosquito surveillance around homes. Compared with people living in Turkey, Mexicans had 14.3 (p 0.0001) times higher odds and Malaysians had 4.0 (p = 0.7030) times the odds of being willing to download a mosquito surveillance app. Compared to urban dwellers, rural dwellers had 1.56 times the odds of being willing to be trained. There is widespread community support for drone use in mosquito surveillance and this community buy-in suggests a potential for success in mosquito surveillance using drones. A successful surveillance and community engagement system may be used to monitor a variety of mosquito spp. Future research should include qualitative interview data to add context to these findings.

Dengue fever is a global life-threatening vector-borne disease which is mainly distributed by the vector Aedes Aegypti mosquito. It is known that the development and survivorship of this vector depends on surrounding climate. The dengue outbreak in Kelantan, Malaysia is alarming. The aim of the study was to compare the fixed effects Negative Binomial GLM and Bayesian Poisson GLMM in prediction of dengue incidence in Kelantan. The data involved daily number of reported dengue cases (1st January 2013–31st December 2017) in ten districts of Kelantan which was collected from Ministry of Health Malaysia. The climate variables, average daily temperature, relative humidity and rainfall (climatic data) were obtained from NASA’s Global Climate Change website, while the population data were from Department of Statistics Malaysia. Statistical modeling results revealed that the fixed effects Negative Binomial GLM failed to fit the daily dengue incidence when serious epidemic occurred. The spatio-temporal Bayesian Poisson GLMM model improved the prediction of dengue incidence. Relative humidity at lag 7 days and 21 days and average temperature at lag 21 days were found to be significant contributing factors of dengue incidence in Kelantan. The findings of the study are significant to respective local authorities in providing vital information for early dengue warning systems in a particular area. This is important for authorities to monitor and reduce dengue incidence in endemic areas and to safeguard the community from dengue outbreak.

The machine learning approach has been widely used in many areas of studies, including the tourism sector. It can offer powerful estimation for prediction. With a growing number of tourism activities, there is a need to predict tourists’ classification for monitoring, decision making, and planning formulation. This paper aims to predict visitors to totally protected areas in Sarawak using machine learning techniques. The prediction model developed would be able to identify significant factors affecting local and foreign visitors to these areas. Several machine learning techniques such as k-NN, Naive Bayes, and Decision Tree were used to predict whether local and foreign visitors’ arrival was high, medium, or low to these totally protected areas in Sarawak, Malaysia. The data of local and foreign visitors’ arrival to eighteen totally protected areas covering national parks, nature reserves, and wildlife centers in Sarawak, Malaysia, from 2015 to 2019 were used in this study. Variables such as the age of the park, distance from the nearest city, types of the park, recreation services availability, natural characteristics availability, and types of connectivity were used in the model. Based on the accuracy measure, precision, and recall, results show Decision Tree (Gain Ratio) exhibited the best prediction performance for both local visitors (accuracy = 80.65) and foreign visitors (accuracy = 84.35%). Distance to the nearest city and size of the park were found to be the most important predictors in predicting the local tourist visitors’ park classification, while for foreign visitors, age, type of park, and the natural characteristics availability were the significant predictors in predicting the foreign tourist visitors’ parks classification. This study exemplifies that machine learning has respectable potential for the prediction of visitors’ data. Future research should consider bagging and boosting algorithms to develop a visitors’ prediction model.

The nonstationary in time series data may be caused by the existence of intervention, outliers, and heteroscedastic effects. The outliers can represent an intervention so that it creates a heteroscedastic process. This research investigates the involvements of these three factors in time series data modelling. It is also reviewed how long the effects of the intervention and outliersfactors will last. The weekly IDR-USD exchange rate in period of May 2015 to April 2020 be evaluated. It is obtained that ARIMA model with the intervention factor gives the best re-estimation result, with smallest average of errors squared. Meanwhile for prediction, the heteroscedastic effect combined with outlier factors gives better results with the lowest percentage of errors. One of the phenomenal interventions in this data is the Covid-19 pandemic, which was started in Indonesia on March 2020. It is found that the effect of the intervention lasts less than five months and the prediction shows that the volatility of IDR-USD exchange rate starts to decline. This shows the stability of the process is starting to be maintained.

COVID-19, CoronaVirus Disease – 2019, belongs to the genus of Coronaviridae. COVID-19 is no longer pandemic but rather endemic with the number of deaths around the world of more than 3,166,516 cases. This reality has placed a massive burden on limited healthcare systems. Thus, many researchers try to develop a prediction model to further understand this phenomenon. One of the recent methods used is machine learning models that learn from the historical data and make predictions about the events. These data mining techniques have been used to predict the number of confirmed cases of COVID-19. This paper investigated the variability of the effect size on the correlation performance of machine learning models in predicting confirmed cases of COVID-19 using meta-analysis. It explored the correlation between actual and predicted COVID-19 cases from different Neural Network machine learning models by means of estimated variance, chi-square heterogeneity (Q), heterogeneity index (I2) and random effect model. The results gave a good summary effect of 95% confidence interval. Based on chi-square heterogeneity (Q) and heterogeneity index (I2), it was found that the correlations were heterogeneous among the studies. The 95% confidence interval of effect summary also supported the difference in correlation between actual and predicted number of confirmed COVID-19 cases among the studies. There was no evidence of publication bias based on funnel plot and Egger and Begg’s test. Hence, findings from this study provide evidence of good prediction performance from the Neural Network model based on a combination of studies that can later serve in the prediction of COVID-19 confirmed cases.

Air pollution is a well-known issue for all countries, including Malaysia. It has been stated that particulate matter that less than 2.5mm known as PM2.5 has a greater effect on health as the smaller particulate size can penetrate deep into the respiratory system and affect the cardiovascular system significantly. Therefore, it is necessary to estimate the concentration of PM2.5 for haze precautions. This study characterizes the pattern of PM2.5 concentrations involving seven stations including Alor Setar, Shah Alam, Pasir Gudang, Ipoh, Kuantan, Kuala Terengganu and Miri with seven indicator parameters (Carbon Monoxide, Ozone, Sulphur Dioxide, Nitrogen Dioxide, Humidity, Temperature and Wind Speed). PM2.5 concentrations were predicted for each station using Multiple Linear Regression (MLR) and Artificial Neural Networks (ANN). Descriptive and trend analysis using Mann-Kandell Trend analysis was used to describe the haze characteristics and identify significant trends in the haze selected locations in Malaysia. MLR and ANN were fitted for the data. The performance of both prediction models was compared based on R2 and Mean Square Error (MSE). The results show ANN performed better than MLR with a high value of coefficient determination (R2) and low error measure. The ANN model was used to predict the occurrence of haze for the next day in the Air Quality Index (API).

According to the World Health Organization (WHO), approximately 2 billion people worldwide use drinking water sources that are contaminated with faeces. This is a serious issue since contaminated water may lead to certain waterborne diseases such as cholera, hepatitis A, dysentery, jaundice, and typhoid fever. Therefore, many researchers around the world are interested in studying the water quality. One of the most commonly used approaches is by using machine learning. Machine learning approach has grabbed the interest of many researchers since the last several years due to its power to compute complicated mathematical computations on big data analysis. Therefore, this study explored the correlation between different water quality parameters and Water Quality Index (WQI) in water quality studies that used machine learning by using a meta-analysis approach. This study used estimated variance, heterogeneity index, Chi-squared heterogeneity test and the random effects model. Based on the selected articles, pH, dissolved oxygen (DO) and biochemical oxygen demand (BOD) are the parameters commonly used in water quality studies which use a machine learning approach. This study found that pH is the best chemical factor which greatly affects the Water Quality Index since it has the highest mean correlation and lowest estimated variance due to sampling error. The result showed that the correlation between pH and WQI are heterogeneous across studies based on the Chi-squared of heterogeneity, Q and heterogeneity index, I2 value. The 95% confidence interval of effect summary supports the findings that the correlation of pH is different among the studies. This study also found that there is no evidence of publication bias using Egger and Begg’s test. Therefore, in order to ensure good water quality supply, the local authorities and government agencies should give more attention to this parameter since pH of water plays an important role in determining the water quality status.

Based on data of visitors to national parks, nature reserves and wildlife sanctuaries in Sarawak, this study’s objective is to use the spatial and temporal analysis to describe the underlying trend and temporal pattern of local and foreign visitors and ultimately infer the temporal distribution of visitors to 18 different TPAs. The second aim of the study is to cluster the visitors according to the location of TPAs using Wards hierarchical clustering method. By comparing average monthly visitors’ count, we observed that the average number of monthly visitors significantly reflects the distribution concentration of visitors based on the spatial map. Findings indicate that the monthly distributions of local and foreign visitors differ according to different TPAs. The spatial and temporal analysis found that local visitors’ arrival is high at the end of the year while foreign visitors showed significant arrival during the months of July, August and September. The Wards minimum variance method was able to cluster TPAs local and foreign visitors into very high, high, medium and low visitor area. This study provides additional information that could contribute to identifying the periods of highest visitor pressure, design measures to manage the concentration of visitors and improve the overall visitors’ experience. The findings of the study are also important to respective local authorities in providing information for planning and monitoring tourism in TPAs. Consecutively, this will ensure sustainability of TPAs resources while protecting their biodiversity.

Objective To examine the impact of haze in the reduction of peak expiratory flow rate (PEFR) reading and identify the risk factors affecting respiratory function due to haze. Methods This study was conducted during haze period among secondary school students in Kota Bharu. We analyzed data on a total of 126 secondary school children measuring the respiratory health and symptoms in October 2015 using standardized questionnaire and PEFR measurement. Clinical characteristics on the risk factor and prevalence of haze effect were explored. Chi -square test and independent sample t -test was used to investigate the relationship between risk factors and haze effect and logistic regression analysis for the odds of having haze effect. Results The findings revealed a significant reduction in PEFR reading of more than 15% from the expected PEFR values. It was also noted that the children with headache, cough, mucus and sore throat respiratory symptoms had consistently higher rates of respiratory illness of having haze effect compared to those who did not. Conclusions Student with haze effect documented much higher symptoms during haze especially female students. Symptoms such as headache, wheezing and mucus were noted among the normal secondary school children in Kota Bharu.

The issue of graduates’ employability has become a major concern in many countries as the number of graduates keep increasing over the years. Fresh graduates are experiencing difficulties in finding meaningful employment as the growth in hiring has slowed in the recent period. With the impact of fourth industrial revolution, today graduates face a world transformed by technology. The role of universities in higher education become more challenging in shaping more digital technology graduates with the right skills and knowledge. Problems can be seen at both management level of higher education in producing graduates who can meet the needs of industry while industry is also facing the problem of finding skilled graduates who suit their needs. To examine the work skills demanded from Industry of Malaysian, this study proposed a text mining approach from 465 unstructured data of analyst job vacancies in classifying job titles and extracting vacancy and skills from job descriptions. The methodology involved are text retrieval, text pre-processing, word cloud corpus generation and text clustering. The findings indicate that the selected job posts emphasize more on vacancies information rather than skills needed which further shows the gap existence between the skills demand and supply.

The response of students to learning is a key element for measuring the extent to which the quality of education is perceived in online learning. Analysing the nature of students' feedback is important in improving student skills and shaping learning behaviour in online learning. Hence, this study aimed to investigate the nature of the feedback itself. We conducted an analysis on students' feedback in web-based learning for two selected courses which were exam based and non-exam based courses. As a result, the feedback received was then categorised using the thematic analysis according to different nature of feedbacks. Next, regression analysis was carried out on the count of discovered nature of feedback to determine its relationship and usefulness as evaluated by the instructor. This study disclosed four disparate natures of feedbacks based on the responses and has significant positive correlation between the use of feedback.

This preliminary study aims to investigate the awareness of the adoption of cloud computing among the academicians in tertiary education in Malaysia. Besides, this study also want to explore the possible challenges faced by the academician while adopting this new technology. The pilot study was done on 40 lecturers in Universiti Teknologi MARA Kampus Kota Bharu (UiTMKB) by using self administered questionnaire. The results found that almost half (40 percent) were not aware on the existing of cloud computing in teaching and learning (T&L) process. The challenges confronting the adoption of cloud computing are data insecurity, data insecurity, unsolicited advertisement, lock-in, reluctance to eliminate staff positions, privacy concerns, reliability challenge, regulatory compliance concerns/user control and institutional culture/resistance to change in technology. This possible challenges can be factorized in two major factors which were security and dependency factor and user control and mentality factor.

Non-fatal road accident injury has become a great concern as it is associated with injury and sometimes leads to the disability of the victims. Hence, this study aims to develop a model that explains the factors that contribute to non-fatal road accident injury severity. A sample data of 350 non-fatal road accident cases of the year 2016 were obtained from Kota Bharu District Police Headquarters, Kelantan. The explanatory variables include road geometry, collision type, accident time, accident causes, vehicle type, age, airbag, and gender. The predictive data mining techniques of decision tree model and multinomial logistic regression were used to model non-fatal road accident injury severity. Based on accuracy rate, decision tree with CART algorithm was found to be more accurate as compared to the logistic regression model. The factors that significantly contribute to non-fatal traffic crashes injury severity are accident cause, road geometry, vehicle type, age and collision type.

Bus driver scheduling system is one of the important aspects to be considered in providing the optimum level of workers efficiency aim to reduce accident occurrence. Lack of systematic and organized scheduling system is one of the issues concerned. This study is carried out to build a model for the problem of scheduling the bus drivers using 0-1 Goal Programming (GP) method. Adapting the case of Transnasional bus driver, the GP model is developed based on problem solving technique by assigning the priority level which takes into account the company policies and drivers' preferences. The solution of the model provides the bus drivers a scheduling system that satisfies all the constraints and is fair in terms of work load distribution as well as uniformed allowances.

The application of Poisson and Negative Binomial models has been widely used in modeling road accident count. However, several restrictions on the data have been highlighted in the use of such model. Among which are the assumption of variance and mean to be equal, no serial correlation exist and the effect of unmeasured variables that may affect the dependent variable or so called the unobserved heterogeneity. Hence, an appropriate solution to this issue is to treat the data in a form time series and cross section data panel using panel model approach. This study analyzes the number of road accidents occurrences using time series cross sectional data for 14 states in Malaysia. The random effects negative binomial (RENB) model and the cross-sectional negative binomial (NB) models are examined. The models were developed to identify the contributing factors that affect the number of road accidents in Malaysia. We examine various factors associated with road accidents occurrence that includes the registered vehicle in the state, the amount of rainfall, the number of rainy day, time trend and the monthly effect of seasonality. Various model specifications were estimated. The specification comparisons indicate benefit from using the NB model when spatial and temporal effects are unobserved. While the RENB model appears to be more superior in the present case of accident count when incorporating temporal and cross sectional variations in which offers advantages in model flexibility.

A number of methods have been proposed for dealing with road accident death model. This paper uses econometric regression models to develop the road accident death model. By using this approach, this paper attempts to establish a statistical model to describe the relationship between the total road accident deaths and a range ofexplanatory macroeconomic variables. The macroeconomic factors used in the model include population, the number of registered vehicles, road length, technique of data coverage, system of data recording and Gross Domestic Product. The results suggest that the POp, ROADL, VEH and DR do not have any impact on road accident deaths. In contrast, the GDP and Technique of data Coverage were found to be highly significant (P < 0.05) in explaining the road accident deaths.