Reported fatal accidents in Malaysia construction industry contributed the highest number among other industries since 2009 until 2018. These statistics are alarming and elucidate the need to conduct scientific research to find the solution to solve the problem. Safety and health issues remain critical to the construction industry due to its working environment and the complexity of working practices. To prevent an accident, hazard identification is essential to construction safety management because unidentified hazards present the most unmanageable risks. Therefore, this research is conducted to help in reducing occupational accidents in the construction industry, and it can be considered worthy. This study aims to understand the triggering events, and their factors leading to fatal accidents are of significant input. The obtained data were analyzed using a statistical analysis program. Data collections of 139 numbers of respondents have been conducted in six different construction sites in east Malaysia. The result showed that there is a correlation between worker's perspective on safety management in the construction site and accident that can lead to a more fatality rate compared to other sectors. The corresponding recommendations are ultimately put forward as fatal accident prevention in construction activities. The patterns found in this paper can contribute valuable direction for formulating accident prevention strategies. In future research, reporting of an accident and near misses are essential to be further explored.

Advancement of high-throughput technologies in omics studies had produced large amount of information that enables integrated analysis of complex diseases. Complex diseases such as cancer are often caused by a series of interactions that involve multiple biological mechanisms. Integration of multi-omics data allows more advanced analysis using features from various aspects of biology. However, analysing cancer multi-omics data on a large scale could be challenging due to the high dimensionality of the data. The recent development of advanced computational algorithms, especially deep learning, had sparkednumerous efforts in applying these algorithms in multi-omics studies. This study aims to investigate how deep learning algorithms, namely stacked denoising autoencoder (SDAE) and variational autoencoder (VAE) can be used in cancer classification using multi-omics data. Moreover, this study also investigates the impact of feature selection in multi-omics analysis through the implementation of an embedded feature selection. The multi-omics data used in this study includes genomics, methylomics, transcriptomics and clinical data for a case study of lung squamous cell carcinoma. The classification performance has beencompared and discussed in terms of the effectiveness of different models and the impact of feature selection. Results showed that VAE outperforms SDAE with 91.86% accuracy, 22.73% specificity and 0.21% Matthews Correlation Coefficient (MCC).

Protease is a proteolytic enzyme that hydrolyzes the amino acid where the cleavage only occurs at specific sites of the amino acid substrate.  By discovering the nick site, the prediction on the function of proteases can be identified and enable humans to control the protein's hydrolysis by their corresponding protease. It is very contributed to controlling protein production especially viral protein. The experts may alter the production of viral protein by reducing the viral proteases to undergo proteolysis. With the rise of computational methods in this era, deep learning is becoming more famous and applied in every field of study, including the biological area. Conventional techniques such as mass spectrometry and two-dimensional gel electrophoresis are being replaced by computational methods due to time-consuming. Thus, this study improves the deep learning algorithm by proposing the Hybrid model of Random Forest + Deep Neural Network (Hybrid RF+DNN) to classify nick sites. The classification in this study is compared with the other machine learning algorithms such as Random Forest (RF), Support Vector Machine (SVM),  and Deep Neural Network (DNN). The proposed method is believed to enhance the classification results in identifying the positive and negative nick sites. The RF is a feature-selector that gathers the most important feature before entering the DNN classifier. This approach reduces the data dimensionality and speeds up the execution time of the training process. The performance of the models was measured by confusion matrix, specificity, sensitivity, etc. However, the proposed method is not the best performer among the mentioned classifiers from the result. The proposed method may become the best performer as the parameter tuning is done more precisely, even after the feature selection by the RF algorithm. Thus, the proposed method with the enhancement appears to be an alternative to the researcher discovering nick site.

The challenge in classifying cancer may lead to inaccurate classification of cancers, especially sarcoma cancer since it consists of rare types of cancer. It is hard for the clinician to confirm the patient's condition because an accurate diagnosis can only be made by the specialist pathology.  Therefore, instead of a single omics is used to identify the disease marker, an approach of integrating these omics to represent multi-omics brings more advantages in detecting and presenting the phenotype of the cancers. Nowadays, the advancement of computational models especially deep learning offered promising approaches in solving high-level omics of data with faster processing speed. Hence, the purpose of this study is to classify cancer and non-cancerous patients using Stacked Denoising Autoencoder (SDAE) and One-dimensional Convolutional Neural Network (1D CNN) to evaluate which algorithm classifies better using high correlated multi-omics data. The study employed both computational models to fit multi-omics dataset. Sarcoma omics datasets used in this study was obtained from the Multi-Omics Cancer Benchmark TCGA Pre-processed Data of ACGT Ron Shamir Lab repository. From the results, the accuracy obtained for the SDAE was 50.93% and 52.78% for the 1D CNN. The result show 1D CNN model outperformed SDAE in classifying sarcoma cancer.

Attention Deficit Hyperactivity Disorder (ADHD) categorize as one of the typical neurodevelopmental and mental disorders. Over the years, researchers have identified ADHD as a complicated disorder since it is not directly tested with a standard medical test such as a blood or urine test on the early-stage diagnosis. Apart from the physical symptoms of ADHD, clinical data of ADHD patients show that most of them have learning problems. Therefore, functional Magnetic Resonance Imaging (fMRI) is considered the most suitable method to determine functional activity in the brain region to understand brain disorders of ADHD. One of the ways to diagnose ADHD is by using deep learning techniques, which can increase the accuracy of predicting ADHD using the fMRI dataset. Past attempts of classifying ADHD based on functional connectivity coefficient using the Deep Neural Network (DNN) result in 95% accuracy. As Variational Autoencoder (VAE) is the most popular in extracting high-level data, this model is applied in this study. This study aims to enhance the performance of VAE to increase the accuracy in classifying ADHD using fMRI data based on functional connectivity analysis. The preprocessed fMRI dataset is used for decomposition to find the region of interest (ROI), followed by Independent Component Analysis (ICA) that calculates the correlation between brain regions and creates functional connectivity matrices for each subject. As a result, the VAE model achieved an accuracy of 75% on classifying ADHD.