Search

Forum and social media especially Facebook, Instagram and Twitter play a vital role in the communication and become the pivotal role in expressing feelings and opinions toward certain posts and news through leaving a comment and review. The comment and review can be positive, neutral, or negative. There have been several longitudinal studies involving online comments showed that one of five comments, 20 percent of comments are toxic comments. Toxic comments are those negative comments that can greatly affect a person emotionally which might cause people to suffer from mental problems. Hence, toxic comments detection is fast becoming a key instrument in protecting social media platforms' users from cyberbullying. This paper seeks to remedy this issue by proposing and comparing different artificial intelligence models to classify the toxic online comments. Thus, six different models with various algorithms which are CNN, RNN, Hybrid-NN, Linear SVC, multinomial naive bayes and logistic regression will be trained with a pre-processed dataset and evaluated with cross validation and ROC AUC. These models are then compared by observing the accuracy of each model and result shows that Hybrid-Neural network, combination between RNN and CNN score the top of the chart with accuracy of 0.9778 which suggest the best model to classify the online comments. [ABSTRACT FROM AUTHOR]

This paper presents the development of a deep convolutional neural network (DCNN) for human detection when using down sampled images. The possible applications for a reliable and accurate human detection DCNN when the image is of low resolution is limitless. The most common application of such DCNN in today's modern society is in the advanced reconnaissance via unmanned aerial vehicles for detecting the presence of a human being during disasters and emergency crisis. While there are traditional methods such as vision detection / recognition systems, DCNN is superior as it is scalable to large data sets such as images, is less complex and easy to setup, and has a predictable output. The DCNN presented in this paper is initially optimized with eight hidden layers using input images of 30 × 30 resolution, having a low loss of 0.1525 and a high accuracy of 0.9444. The model is then further trained and validated with images of higher resolutions of 50 × 50 and 100 × 100. The later evaluation demonstrated that going beyond image resolutions of 50 × 50, the increase in computation time does not justify the minimal improvement in model loss and accuracy. Finally, the study demonstrates that the model performance can be further improved without any further optimization by simply retraining the model for a second time. This resulted in a DCNN model with eight hidden layers having a model loss of 0.0902 and accuracy of 0.9827 when using input images of 50 × 50 resolution. [ABSTRACT FROM AUTHOR]

Application of solar energy in a portable solar charger is a sustainable solution especially at locations that are far away from the electricity grid. This study presents a systematic approach in modeling and simulation of solar photovoltaic (PV) array with partial shading conditions. Specifically, the partial shading conditions are based on portable solar charger applications that are modular and can be carried anywhere on the top of a backpack. The proposed method is reasonably accurate and very useful as a preliminary study before designing the interconnection of cells within a solar PV module. Using a mannequin to represent a person carrying a portable solar charger on the top of a backpack, eight scenarios of partial shading is considered assuming that the person is facing eight different directions relative to true north. From there, partial shading patterns are applied in MATLAB Simulink modeling and simulation of the photovoltaic array. The results are verified using real outdoor irradiance measurements at Ipoh, Malaysia. In addition, the maximum output power is determined and compared with simulated output power using MATLAB. It was found that the accuracy of the results is in the range of 94.41% to 97.08%. It is envisioned that the results are important in designing and optimizing solar portable charger that are integrated with camping backpack for anyone that is travelling on foot. [ABSTRACT FROM AUTHOR]

A numerical study is presented on different solar cell configurations of a photovoltaic module based on the impact of non-uniform solar irradiance. MATLAB is used to simulate the module using electrical parameters such as open-circuit voltage, Voc, short-circuit current, Isc, maximum peak voltage, Vmp, maximum peak power, Pmp and efficiency, η. The module consists of 60 solar cells, and it is simulated based on series, parallel, total-cross-tied (TCT) and series-parallel (SP) interconnection. The results of the simulation for uniform irradiance were compared with a commercial data sheet which yielded an agreement within ±1.7%, on average. Based on the validated model, the research was extended to non-uniform irradiance, and the results show that cell configuration 2 × 30, 3 × 30 and 6 × 10 produced the highest output power using TCT configuration. On the other hand, SP configuration produced the highest output power at 3 × 20 configuration. All the configurations mentioned achieved an efficiency of 13.1%. An optimal cell configuration of 6 × 10 is selected based on the highest normalized values of Voc and Isc. [ABSTRACT FROM AUTHOR]

One of the key parameters of a country's socio-economic achievement is the construction of roads and highways. The technology for road construction has always been improved particularly in terms of quality and environmental aspects, for which the idea of using plastic emerged recently. Plastic roads are an auspicious strategy for lowering the quantity of land filled plastic, minimizing the dependency on depleted bituminous materials, and enhancing asphalt pavement performance. Assessing the environmental and human health aspects of the modified asphalt mixtures estimates the safety level of the product when exposed to the ecosystem. The objective of this study is to assess the preliminary environmental and human health hazards of our polypropylene (PP), high-density polyethylene (HDPE), and low-density polyethylene (LDPE) microplastics modified asphalt mixtures using fluorescence microscopy technique and polymer risk index. The mixtures were prepared by wet mixing the asphalt binder with different proportions of powdered microplastic polymers. The asphalt mixture that contained 100% HDPE had the highest crystal formation. Combining at least 5g of PP, reduced the risk index of the asphalt mixture. All the plastic polymers used in our study are ranked as the least environmentally hazardous chemicals. [ABSTRACT FROM AUTHOR]

Torrefaction is one of the pre-treatment methods used to improve properties of biomass in order to be utilized as feedstock for pyrolysis and gasification processes. Prior knowledge of kinetic mechanism is significant before biomass can be utilized as feedstock for energy application. Thermogravimetric analysis (TGA) is an essential technique for kinetic study of raw and torrefied biomass. In this work, the effect of torrefaction on two different parts of oil palm frond (petiole and leaflets) on pyrolysis behaviour and kinetic parameter of the process were investigated. The results of this work showed removal of moisture and low molecular volatile matters contributes to the mass loss for both petiole and leaflet samples in stage Ⅰ. Coats-Redfern (CR) model is used in estimating activation energy (Ea) and pre-exponential factor (A) for stage Ⅱ (decomposition of hemicellulose and cellulose) and stage Ⅲ (decomposition of lignin). The regression with R2 values higher than 0.88 are obtained indicating the reliability of estimated kinetic parameters. The Ea and A values of petiole and leaflets show increment trends in both stage Ⅱ and stage Ⅲ. As comparison, torrefied petiole is the most suitable feedstock for pyrolysis process due to the highest HHV (23.55 MJ/kg) and produce highest char yield (41.21 wt.%). [ABSTRACT FROM AUTHOR]

Organic Rankine cycle (ORC) is one of the solutions for recovering waste heat into useful power output. Low-grade waste heat from various sources can be converted into electricity using ORC in many different aspects. The typical ORC systems in the market often require large amount of waste heat as the waste heat sources. Although the thermal efficiency for marketed ORC is reasonable, it will not be practical in the condition of limited available energy from small-scale waste heat or low temperature heat sources. Hence, this paper presents the numerical simulation of small-scaled ORC using design model developed in MATLAB to study the thermal efficiency of various working fluids in limited working conditions. The fluid properties were obtained from CoolProp library. Result shows that R123 has the highest thermal efficiency of 7.25% with net power output of 4.09kW for this small-scaled ORC. The ranking of refrigerants as working fluids used in the designated working condition is also presented. [ABSTRACT FROM AUTHOR]

Blackjack is a card game that usually played in casino where one or several players bet against a dealer. Each player and the dealer are given a pair of cards where both cards received by the player are facing up (known as the upcard) and the dealer has one card facing up and another facing down (known as the hole card). The rule is simple, the players must accumulate points to beat the dealer. The points of each card are the number that appears on the card, with cards like J, Q and K are counted 10 points. An A can be counted as 1 or 10 points at the player's preference. Additional cards (maximum 5 cards) can be drawn be players and the dealer and both parties must not exceed 21 points (busting). This game can be played instinctively with luck but mathematicians had found a systematic way to increase the player's advantage in Blackjack; which is known as card counting. There are many card counting techniques such as Hi-Low count, Hi-Opt II, Red 7, etc. In this study, we will determine whether the Hi-Low count technique will affect the winning probabilities of a player as well as the profit gained by a player in Blackjack. We simulate Blackjack for 10, 000 and 100, 000 trials in MATLAB 2018a using the Hi-Low count and compare the result against the trials without card counting. Furthermore, we will also vary the number of decks used in the trials as this to see how it affects the efficiency of card counting. [ABSTRACT FROM AUTHOR]

This study focuses on the empty container management problem in a deficit regional area. Deficit area is the area having more export activities than the import activities, which always have a shortage of empty container. This environment has challenged the trading companies in the decision making in distributing the empty containers. A simulation model that fit to the environment is developed. Besides, a simple heuristic algorithm with some hard and soft constraints consideration are proposed to plan the logistic of empty container supply. Then, the feasible route with the minimum cost will be determined by applying the proposed heuristic algorithm. The heuristic algorithm can be divided into three main phases which are data sorting, data assigning and time window updating. [ABSTRACT FROM AUTHOR]

We show that in the current practice of multivariate process variability monitoring, the reliability of Shewhart-type control charts cannot be measured except when the sub-group size n tends to infinity. However, the requirement of large n is meaningless not only in manufacturing industry where n is small but also in service industry where n is moderate. In this paper, we introduce a new definition of control limits in the two most appreciated control charts in the literature, i.e., the improved generalized variance chart (IGV-chart) and vector variance chart (VV-chart). With the new definition of control limits, the reliability of the control charts can be determined. Some important properties of new control limits will be derived and the computational technique of probability of false alarm will be delivered. [ABSTRACT FROM AUTHOR]

Natural sand does always contain significant amounts of fine particles, known as sand matrix soils. The development of critical state for sand and clay are both well established. However, limited researchers focused on the sand matrix soils. This paper discusses the effect of fines content on the critical state parameters of sand matrix soils. Two types of plastic fines with different plasticity (kaolin and bentonite) were mixed into poorly graded Johor clean sand at different percentages to form the sand matrix soils. The critical state was obtained from the strain controlled triaxial compression test. All tested soils had been isotropically consolidated to four different effective confining pressures and sheared under consolidated undrained condition until the critical state was ascertained. The critical state line was plotted for each sand matrix soils on space and compression spaces. Based on linear regression analysis, the critical state parameters, M, λ and Γ were obtained and the ranges are from 0.807 to 1.102, 0.062 to 0.160 and 1.732 to 2.431, respectively. The trend of the critical state lines of sand matrix soils in both stress and compression space coincides with the concept of threshold fines content. The threshold fines content identified for sand-kaolin mixtures is 25 % kaolin while it is 20 % bentonite for sandbentonite mixtures. [ABSTRACT FROM AUTHOR]