
Md. Atikur Rahman
Lecturer
At a glance
Biography
I am passionate about teaching, research, and continuous learning. My research interests include hyperspectral image processing, deep learning, computer vision, and artificial intelligence. I enjoy exploring emerging technologies, sharing knowledge, and applying innovative ideas to solve real-world problems.
Education
| Degree | Institution | Year | Result |
|---|---|---|---|
| M.Sc.Engg. in Computer Science and Engineering | Rajshahi University of Engineering & Technology | Pursuing | |
| B.Sc. in Computer Science and Engineering | Pabna University of Science and Technology | 2022 |
Research Interests
Publications
Journal Publications
Credit card companies must be able to identify fraudulent credit card transactions so that clients are not charged for items they did not purchase. Previously, many machine learning approaches and classifiers were used to detect fraudulent transactions. However, because fraud patterns are always changing, it is becoming increasingly vital to investigate new frauds and develop the model based on the new patterns. The purpose of this research is to create a machine learning classifier that not only detects fraud but also detects legitimate transactions.
We consider a deterministic optimal control approach with cost-effectiveness analysis for the diphtheria outbreak in the Rohingya refugee camp in Bangladesh. A deterministic epidemic dynamical model for diphtheria outbreaks has been developed with three optimal controls: vaccination, latent, and infectious treatment.
One exciting area within computer vision is classifying human activities, which has diverse applications like medical informatics, human-computer interaction, surveillance, and task monitoring systems. In the healthcare field, understanding and classifying patients’ activities is crucial for providing doctors with essential information for medication reactions and diagnosis.
Conference Publications
Hyperspectral image (HSI) classification requires efficient utilization of spectral and spatial data in order to reach proper identification of ground materials. The patch-based data splitting is employed in most currently used methods, and it may result in the overlap of training and testing patches and produces overly optimistic and unrealistic results. To solve this problem, this paper proposes a block-based spatial splitting approach that trains and tests the image into non-overlapping blocks so that no data is leaked and fair evaluation is conducted.
Hyperspectral Images (HSI) consist of hundreds of consecutive narrow spectral bands and have widely been used in ground cover analysis of remote sensing, agriculture, and environmental observation. However, due to their high dimensionality and class imbalance, it is challenging to yield discriminative features from multiple classes. Convolutional Neural Networks (CNNs) have only been able to learn local features, while transformers can learn global dependencies but typically in the form of higher model complexity. To get around these problems, we propose an effective Spectral-Spatial Fusion Network (SSFN) which combines local and global features.
Professional Experience
Workshops & Seminars
Seminars & Talks
Software Engineering and Cloud Computing
Mobile Game and Application
Contact
Send a message to Md. Atikur Rahman through the department office. Direct email and phone details are listed above.