Welcome to AIBIG Research Group

Identification of Multi-Omics Features for Cancer Classification Using Improved Stacked Denoising Autoencoders.

-Developed Using Deep Learning Method to identify significant multi-omics features that relates to the phenotypes of interest.

-contributes to an improvement in identifying the informative features from the multi-omics data for cancer classification which can be applied in the health sector (Healthcare is one of the 12 important NKEA) by biologists and doctors in Malaysia for cancer diagnosis. Furthermore, the proposed improved algorithm could be applied to other big data namely metabolomics data and real-time PCR data, with huge potential in establishing collaborations with the private sectors in areas such as pharmaceutical and clinical research (beneficial to drugs discovery, which supports the aims of National Biotechnology Policy objective ).

Multi Objective Optimization of Building Information Modelling

-An Intelligent Decision Making (DM) tool

– Developed Using AI techniques and integrated with Building Information Modelling (BIM) to design and analyze optimized and sustained green building

-The process includes specifications on combination of building orientation, material selection and design analysis

BIOMARKERS DISCOVERY USING IMPROVED PENALIZED MACHINE LEARNING ALGORITHM

– Developing intelligent method using penalized machine learning to identify biomarkers from topological information extracted through integration of gene expression, biological pathways and interactions.

– Aid to provide better and efficient discovery of biomarkers through the data and give useful analytic and mechanistic insights in disease prognosis, drugs design and treatment decision.