Md Fahim Sikder is currently pursuing his Ph.D. under the guidance of Professor Fredrik Heintz at the Reasoning and Learning (ReaL) Lab, IDA, Linköping University, Sweden. His research focuses on creating Generative Models for Time-Series and Fair Data Generation. Before this, Fahim served as a Lecturer in the Computer Science and Engineering department at the Institute of Science, Trade, and Technology (ISTT). He also took on the roles of Coordinator of the HEAP Programming Club and Coach of the ACM ICPC team at ISTT.
Fahim’s research interests include Artificial Intelligence, Generative Models, Trustworthy AI.
Md Fahim Sikder conducted several workshops and seminars, including a Workshop on Latex and a Week-long training course on Python (Beginning to Advance including Machine Learning). He participated in several National and International Contests. He achieved many titles, including “Champion at International Contest on Programming and System Development (ICPSD), 2014”, and “Champion at NASA SPACE APPS CHALLENGE 2016” in Rajshahi Region, Bangladesh.
Ph.D. in Computer Science, Ongoing
Linköping University, Sweden
Master of Science in Computer Science, 2018
Jahangirnagar University, Bangladesh
Bachelor of Science (Engineering) in Computer Science & Engineering, 2016
Bangabandhu Sheikh Mujibur Rahman Science & Technology University, Gopalganj, Bangladesh
Higher Secondary Certificate Examination, 2012
Khilgaon Government High School, Bangladesh
Secondary School Certificate Examination, 2010
Khilgaon Government High School, Bangladesh
Department of Computer and Information Science (IDA)
Responsibilities include:
Department of Computer Science & Engineering (CSE)
Responsibilities include:
Synthetic data generation offers a promising solution to enhance the usefulness of Electronic Healthcare Records (EHR) by generating realistic de-identified data. However, the existing literature primarily focuses on the quality of synthetic health data, neglecting the crucial aspect of fairness in downstream predictions. Consequently, models trained on synthetic EHR have faced criticism for producing biased outcomes in target tasks. These biases can arise from either spurious correlations between features or the failure of models to accurately represent sub-groups. To address these concerns, we present Bias-transforming Generative Adversarial Networks (Bt-GAN), a GAN-based synthetic data generator specifically designed for the healthcare domain. In order to tackle spurious correlations (i), we propose an information-constrained Data Generation Process (DGP) that enables the generator to learn a fair deterministic transformation based on a well-defined notion of algorithmic fairness. To overcome the challenge of capturing exact sub-group representations (ii), we incentivize the generator to preserve sub-group densities through score-based weighted sampling. This approach compels the generator to learn from underrepresented regions of the data manifold. To evaluate the effectiveness of our proposed method, we conduct extensive experiments using the Medical Information Mart for Intensive Care (MIMIC-III) database. Our results demonstrate that Bt-GAN achieves state-of-the-art accuracy while significantly improving fairness and minimizing bias amplification. Furthermore, we perform an in-depth explainability analysis to provide additional evidence supporting the validity of our study. In conclusion, our research introduces a novel and professional approach to addressing the limitations of synthetic data generation in the healthcare domain. By incorporating fairness considerations and leveraging advanced techniques such as GANs, we pave the way for more reliable and unbiased predictions in healthcare applications.