Ng Keng Yap

Ng Keng Yap

Auto-generated profile, based on publicly available information

Ng Keng Yap

Institute for Mathematical Research (INSPEM), Universiti Putra Malaysia Seri Kembangan, Malaysia

Research Focus & Metrics

Academic Career Length

6Years (2021-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
19 May 2021 in IEEE Access
Verified
Journal Article
Open access
FNFaranak Nejati
Faranak Nejati
AGAbdul Ghani
Abdul Ghani
NYNg Keng Yap
Ng Keng Yap
AJAzmi Bin Jafaar
Azmi Bin Jafaar
Component-based software development (CBSD) is an alternative approach to constructing software systems that offers numerous benefits, particularly in decreasing the complexity of system design. However, deploying components into a system is a challenging and error-prone task. Model-checking is one of the reliable methods to systematically analyze the correctness of a system. Its brute-force checking of the system’s state space assists to significantly expand the level of confidence in the system. Nevertheless, model-checking is limited by a critical problem called state space explosion (SSE). To benefit from model-checking, an appropriate method is required to reduce SSE. In the past two decades, a great number of SSE reduction methods have been proposed containing many similarities, dissimilarities, and unclear concepts in some cases. This research, firstly, plans to present a review of SSE handling methods and classify them based on their similarities, principle, and characteristics. Second, it investigates the methods for handling SSE problem in the verification process of CBSD and provides insight into the potential limitations, underlining the key challenges for future research efforts.

5 March 2026 in IEEE Access
Verified
Journal Article
Open access
JJJianqing Jiang
Jianqing Jiang
NHNor Asilah Wati Abdul Hamid
Nor Asilah Wati Abdul Hamid
NYNg Keng Yap
Ng Keng Yap
CCChoo Wei Chong
Choo Wei Chong
The growth of internet lending has created a need for decision frameworks based on models that are both personalized and causally interpretable. Conventional uplift models detect treatment responsiveness without recognizing user heterogeneity, the temporal consistency of user behavior, or the upstream design choices that carry important causal implications. This paper proposes an integrated and reproducible Causal User Profiling (CUP) framework that combines causal inference, uplift modeling, and response-based segmentation within a single pipeline. CUP realizes treatment-effect heterogeneity through a four-type response taxonomy (Persuadable, Sure Thing, Lost Cause, Do-Not-Disturb) and embeds it in a multi-stage pipeline involving hybrid feature selection (Information Value (IV), Causal Forest importance, Population Stability Index (PSI) stability, and Stepwise refinement), stratified clustering with a “C2 replacement strategy,” and meta-learning via both the X-Learner and the Doubly Robust (DR) Learner using Logistic Regression (LR). A component-wise ablation analysis finds that feature selection increases AUUC by 25–30%, C2 clustering by 10–12%, and the DR-Learner + LR by another 5–8%. Overall, the integrated CUP framework yields 45–50% higher AUUC than the baseline (“all features + no clustering + standard learner”) while retaining behaviorally coherent and temporally stable insights. Methodologically, we provide: (i) an end-to-end causal user profiling framework that interoperates profiling, causal estimation, clustering, and uplift evaluation; (ii) a behaviorally and causally consistent response segmentation mechanism grounded in the potential-outcomes model; and (iii) a reproducible experimental design that quantifies pipeline-level uplift gains through systematic ablation. Applied to large-scale internet-lending data, CUP reveals opportunities for treatment-aware personalization, enabling financial institutions to target Persuadables, support Sure Things, and avoid disturbing Do-Not-Disturbs based on causal evidence.

5 September 2025 in IEEE Access
Verified
Journal Article
Open access
ANAmin Arab Najafabadi
Amin Arab Najafabadi
NYNg Keng Yap
Ng Keng Yap
ZAZahra Nazemi Ashani
Zahra Nazemi Ashani
FNFaranak Nejati
Faranak Nejati
Mohamed Abdullahi Abdullahi AliMohamed Abdullahi Abdullahi Ali
Mohamed Abdullahi Abdullahi Ali
Ensemble learning plays a crucial role in deep learning (DL) by improving classification accuracy through the aggregation of multiple classifiers. However, traditional methods, such as bagging, boosting, and stacking, face limitations when combining heterogeneous classifiers trained on unrelated tasks, owing to their reliance on closely aligned objectives. To address these challenges, we propose Adjusted Exponential Scaling (AES), a novel approach that combines diverse classifiers using probabilistic outputs, such as those generated by the Softmax function. AES eliminates the need to modify DL architectures or interfere with their training processes, thereby significantly reducing computational costs while ensuring compatibility with existing systems. We demonstrate AES scalability in simulations, successfully combining the outputs of up to 40 classifiers while maintaining probabilistic integrity. A formal verification using the Z3 Satisfiability Modulo Theories (SMT) solver confirmed its accuracy and reliability. AES is designed for simplicity and adaptability, making it particularly well suited for dynamic applications such as deepfake detection and medical diagnostics, where real-time combination is critical. By providing a scalable and computationally efficient solution, AES expands the applicability of ensemble learning and establishes a foundation for future extensions to sequential and binary tasks, further enhancing its versatility in machine learning challenges.

5 September 2026 in International Journal of Innovation and Technology Management
Verified
Journal Article
Closed access
PWPeng Wei Wang
Peng Wei Wang
YWYi Jie Wang
Yi Jie Wang
WCWei Chong Choo
Wei Chong Choo
KNKeng Yap Ng
Keng Yap Ng
RBRan Bi
Ran Bi