Research and Publication Summary

Explore my research and publications:

I am currently an Information Security Auditor and researcher. My current research interests involve Machine Learning, Network Security, and AI in Telecommunication Engineering.

Explore my research and publications...

Additional Information

  • A Survey on Tackling Heterogeneity in Federated Learning: A collaborative work with Kristiania University of Applied Sciences.
  • A Comparative Analysis of Diverse Machine Learning Techniques in Intrusion Detection Datasets (Project Link)
    • Built an interactive user interface demonstrating data preprocessing, standardization, and splicing.
    • Implemented feature selection techniques using regularizers to reduce multicollinearity.
    • Integrated options for classifier selection, ROC-AUC curve analysis, cross-validation, and performance visualization.
  • Identification of Cucumber Leaf Disease: Deep learning-based classification (In progress).
  • Advanced Persistent Threat Detection: Applied unsupervised learning to the CICAPT IoT dataset, addressing imbalanced data challenges [Work under Progress]
  • A novel approach for PCOS Prediction: Symptom-based PCOS prediction using feature selection and Optimized Machine Learning Algorithms Compared evaluation metrics between symptom integrated dataset and only medical assessed dataset.
  • Modeling a 5R Robotic Manipulating System with 5DoF using GUI MATLAB Simulation, 4th ICMIME, Rajshahi
  • Log Analysis and Dashboard Building: Developed statistical models with Python and visualizations using Grafana.
  • Advanced Persistent Threat Detection: Applied unsupervised learning to the CICAPT IoT dataset, addressing imbalanced data challenges [Work under Progress]
  • Undergraduate Thesis: Analysis of the Effect of System and Channel Parameter Variation on Out-of-Band Aided Reconstruction Methods in mmWave and Sub-6 GHz bands.

Modeling of a 5R Robotic Manipulating System with 5DoF using GUI MatLab Simulation

This paper focuses on a planner Robotic Arm System Having 5 Revolute Joints and 5 Degrees of Freedom.

A Comparative Analysis of Diverse Machine Learning Techniques in Intrusion Detection Datasets

A WebApp was based on this Research that Automates the hefty Preprocessing of Data, Lets User choose a Regularizer of Choice and a classifier for that Regularized data to get Classified and Detected By. A Cross Validation Option is currently being Developed.

Identification of Cucumber Leaf Disease: Deep learning-based classification

The major contribution of this paper is the dataset that were created to detect four different diseases of the cucumber- the cash crop of Bangladesh. The models that were trained and evaluated via this novel dataset are CNN, InceptionV3, and EfficientNetB4.The dataset contains four classes of images namely: Angular leaf Spot, Downy Mildew, Powdery Mildew, and good leaf.

A novel approach for PCOS Prediction: Symptom-based PCOS prediction using feature selection and Optimized Machine Learning Algorithms

Compared Assessment of evaluation metrics between symptom integrated dataset and only medical assessed dataset.

HTML5 & CSS3

Frontend and Backend portion of Website Development using PHP and XAMPP.

Machine Learning

Panda, NumPy, Dask, TensorFlow, Tkinter [Small Game Development], PyPlot, Seaborn

DevOps

Docker, Jenkins

Cyber Security

Linux Kernel, Seed Lab, Kali Linux

Contact Me

If you have a Research Position Available at your Lab that involves Machine Learning or Network Engineering or even better both, please do check out my skills and research experience and reach out to me via qmohona@gmail.com