Aizhen Ren | Machine learning | Excellence in Innovation Award

Published on by Engineering Scientist

Prof. Aizhen Ren | Machine learning
| Excellence in Innovation Award

College of Science, Inner Mongolia Agricultural University | China

Prof. Aizhen Ren research work centers on machine learning, deep learning, statistical inference, and their applications in economics, finance, and computational biology. A significant portion of the research contributes to the development and mathematical validation of advanced bootstrap techniques, including the speedy double bootstrap method, which enhances the statistical reliability of phylogenetic tree estimation and provides third-order accurate unbiased p-values. These methods have been applied to evolutionary analyses of horse breeds, supporting biological and genomic investigations with high-precision statistical tools. In the financial domain, the research explores machine-learning-based trend prediction models, such as multiscale bootstrap-corrected random forest voting systems used to forecast stock index movement with improved accuracy and inference reliability. Additional work includes the construction of financial risk early-warning models for listed companies using multiple machine learning approaches, reflecting an interdisciplinary blend of statistics, computing, and economics. Contributions also extend to consumption behavior analysis employing regression-based models, as well as deep learning ensemble frameworks integrating empirical mode decomposition and temporal convolutional networks for time-series prediction tasks. The released R package SDBP operationalizes the novel bootstrap methodology, enabling researchers to compute unbiased p-values efficiently. Overall, the research advances methodological innovation and practical applications across data-intensive scientific domains.

 Profile: Orcid

Featured Publications

Ren, A., Duan, Y., & Liu, J. (2025). Multiscale bootstrap correction for random forest voting: A statistical inference approach to stock index trend prediction. Mathematics, 13(22), 3601. https://doi.org/10.3390/math13223601

Ren, A., Ishida, T., & Akiyama, Y. (2020). Mathematical proof of the third order accuracy of the speedy double bootstrap method. Communications in Statistics – Theory and Methods, 49(16), 3950–3964. https://doi.org/10.1080/03610926.2019.1594295

Ren, A., Ishida, T., & Akiyama, Y. (2013). Assessing statistical reliability of phylogenetic trees via a speedy double bootstrap method. Molecular Phylogenetics and Evolution, 67(2), 429–435. https://doi.org/10.1016/j.ympev.2013.02.011

Jamshaid Ul Rahman | Artificial Intelligence | Editorial Board Member

Published on by Engineering Scientist

Dr. Jamshaid Ul Rahman | Artificial Intelligence
| Editorial Board Member

Abdus Salam School of Mathematical Sciences, GC University Lahore | Pakistan

Dr. Jamshaid Ul Rahman’s research spans artificial intelligence, deep learning, mathematical modeling, and scientific computing, with a strong emphasis on applying advanced neural architectures to complex problems in molecular modeling, epidemiology, biomechanics, and computational physics. His work integrates graph neural networks, wavelet neural networks, periodic neural networks, and deep optimization frameworks to improve prediction accuracy and interpretability in molecular property analysis, virus transmission modeling, and dynamic system simulation. He has contributed significantly to AI-driven analysis of infectious diseases such as Marburg, Ebola, and Hepatitis C through nonlinear modeling, hybrid optimization, and neural simulation techniques. His research also advances computational mechanics, including deep learning-based vibration analysis in tapered beams, micro-electromechanical systems, and nonlinear oscillators. In cheminformatics, he has enhanced molecular property prediction using quantized GNNs, optimized activation functions, and improved training strategies for message-passing networks. His contributions extend to biophysical modeling, where neural networks simulate glycolysis, biochemical oscillators, and biological system dynamics with high fidelity. He has also explored angular softmax variations, Laplacian smoothing, and stochastic gradient methodologies in deep convolutional networks, advancing core theoretical aspects of machine learning. Across more than fifty publications, his work consistently integrates mathematical rigor with cutting-edge AI to address interdisciplinary scientific and engineering challenges.

 Profile:  Google Scholar

Featured Publications

Rahman, J. U., Noureen, I., Mannan, A., & Uwitije, R. (2025). PhyFold: Environment aware physics informed neural network for protein folding dynamics.

Ul Rahamn, J., Iqbal, M. A., Rasool, A., & Uwitije, R. (2025). p-GIN: A graph isomorphism network based on p-Laplacian operator to enhance molecular property prediction. Discover Applied Sciences, 7(11), 1251.

Shoket, N., Ul Rahman, J., & Zafar, N. (2025). Mathematical modeling and computational investigation of the COVID-19 epidemic using wavelet neural networks and coupled optimization. SN Soft Computing, 1–18.

Rahman, J. U., Ali, H., & Rassol, A. (2025). A comprehensive analysis of optimizers in message passing neural networks for molecular property prediction task. Computational Biology and Chemistry, 108556

Prof. Sharmila S P | Computer Engineering | Editorial Board Member

Published on by Engineering Scientist

Prof. Sharmila S P | Computer Engineering
| Editorial Board Member

Siddaganga Institute of Technology Tumakuru | India

Prof. Sharmila S P the research work focuses on advancing cybersecurity through AI-driven, explainable, and resilient detection mechanisms capable of addressing modern, highly obfuscated threats. Central contributions include the development of memory-forensic-based feature extraction techniques that enhance the transparency and interpretability of obfuscated malware detection models, enabling isolated family distinction and reducing false positives. The work explores multi-class classification frameworks for malware analysis, leveraging machine learning paradigms to identify sophisticated adversarial behaviors across diverse threat categories. Additional research investigates Hidden Markov Model–based intrusion detection, employing a randomized Viterbi algorithm to strengthen anomaly recognition in dynamic network environments. Studies on cyber-attack prediction further analyze prevalent forecasting techniques to improve proactive defense capabilities. Complementary research examines Android malware behavior, distributed ledger applications for secure banking operations, and lightweight authentication mechanisms rooted in keystroke dynamics for user verification. With a strong emphasis on AI, machine learning, GNNs, NLP-driven analysis, reverse engineering, and volatile memory forensics, the overall body of work contributes toward building robust, explainable, and scalable cybersecurity systems capable of safeguarding digital infrastructures against evolving threats in cloud environments, embedded systems, mobile platforms, and large-scale networked ecosystems.

 Profile:  Orcid 

Featured Publications

Sharmila, S. P., Gupta, S., Tiwari, A., & Chaudhari, N. S. (2025). Unveiling evasive portable documents with explainable Kolmogorov–Arnold networks resilient to generative adversarial attacks. Applied Soft Computing, 138, 113537. https://doi.org/10.1016/j.asoc.2025.113537

Sharmila, S. P., Gupta, S., Tiwari, A., & Chaudhari, N. S. (2025). Leveraging memory forensic features for explainable obfuscated malware detection with isolated family distinction paradigm. Computers and Electrical Engineering, 121, 110107. https://doi.org/10.1016/j.compeleceng.2025.110107