Other Academic Research and Academic Projects:

1. Building Models for the Classification of the Presence of Heart Diseases 

   • Supervisor: Dr. Md. Rezaul Karim, Professor, Jahangirnagar University & East West University, MSc (Hasselt University), PhD. (Arenberg Doctoral School at KU Leuven, Belgium).

   • Developed and evaluated machine learning models, including Random Forests, Gradient Boosting, and Artificial Neural Networks (ANNs), to classify heart disease presence, focusing on data preparation, dimensionality reduction, and model optimization.

   • The result demonstrated the superior performance of Random Forests and Gradient Boosting in handling non-linear relationships, and highlighting the limitations of ANNs when applying Principal Component Analysis (PCA), providing key insights for optimizing predictive analytics in heart disease diagnosis.

2. Building ML Model to Predict Whether the Tumor being Benign or Malignant (Breast Cancer Dataset)

   • Supervisor: Dr. Md. Rezaul Karim, Professor, Jahangirnagar University & East West University.

   • Constructed robust machine-learning models for accurate tumor classification as either benign or malignant involving six major classifiers: Gaussian Naive  Bayes, Decision Tree, Random Forest, K-Nearest Neighbors, Logistic Regression, and Support  Vector Machine (SVM).

   • The results highlight the Random Forest Classifier as the optimal model, achieving an accuracy of  94.2% and demonstrating superior performance in cross-validation.

3. D-Day Invasion: Development of a comprehensive database capturing the essence of World War II

   • Supervisor: Dr. Syed Akhter Hossain, Dean and Professor of CSE, Daffodil International University, Professor, East West University. Post Doctoral Fellow, LIESP Laboratory, Universite Lyon 2, Lyon, France.

   • Developed a database integrating data on World War II, featuring interactive visualizations to enhance user engagement and understanding.

   • Processed a robust ER model, normalization, schema development, and detailed SQL scripting for data accuracy and performance, providing an accessible platform for students and researchers to explore historical insights.

4. Business Process Modeling for Ready-Made Garments (RMG)

   • Supervisor: Dr. Syed Akhter Hossain, Dean and Professor of CSE, Daffodil International University, Professor, East West University.

   • Developed detailed Event-driven Process Chain (EPC) models to map and analyze key workflows in Utah Group's denim manufacturing, identifying critical bottlenecks and process dependencies.

   • Provided strategic recommendations based on As-Is and To-Be models, including adopting advanced automation technologies and improving real-time inventory management, to streamline operations and enhance efficiency while maintaining high product quality.

5. Comparative Analysis of Multiple Classification Models for Classifying Images from the Fashion MNIST Dataset 

   • Supervisor: Dr. Md. Rezaul Karim, Professor, Jahangirnagar University & East West University. 

   • Constructed and compared robust CNN and ANN models for image classification on the Fashion MNIST dataset, focusing on data preprocessing, model development, and evaluation.

   • The result demonstrates that the CNN model outperformed the ANN model with a higher accuracy (91.07% vs. 88.56%), highlighting the CNN's superior ability to capture complex spatial correlations within low-resolution images.

6. ANN Model for Boston and Tumor Data 

   • Supervisor: Dr. Md. Rezaul Karim, Professor, Jahangirnagar University & East West University.

   • Implemented an Artificial Neural Network (ANN) to predict housing prices and breast cancer classification, focusing on thorough data preprocessing and feature exploration and building an efficient model architecture tailored for each task.

   • For the breast cancer classification model, compared six machine-learning classifiers, including Random Forest, SVM, and Naive Bayes, identifying Random Forest as the optimal model with a 94.2% accuracy, demonstrating superior performance through cross-validation.

7. Optimizing Marketing Campaign Process: IMPROVING THE EFFECTIVENESS AND EFFICIENCY OF MARKETING CAMPAIGNS OF NAGAD

   • Supervisor: Dr. Syed Akhter Hossain, Dean and Professor of CSE, Daffodil International University, Professor, East West University.

   • Identified key bottlenecks in the current marketing campaign processes by analyzing inefficiencies such as unclear objectives, fragmented communication, misaligned resources, and repetitive manual tasks, which hinder the effectiveness of the marketing campaigns.

   • Proposed optimized solutions through a TO-BE process model using techniques like machine learning-driven optimization, simultaneous data management, and early cross-functional collaboration, aimed at improving time efficiency, decision-making, and the overall campaign outcomes

8. Empirical Analysis of the Impact of Foreign Remittances on the Economic Development of Bangladesh 

   • Supervisor: Mr. Kamruzzaman, Additional Director, BBTA, Bangladesh Bank.

   • Conducted an empirical analysis of the impact of foreign remittances on Bangladesh's economic development, utilizing regression analyses and macroeconomic data from the past decade.

   • Findings reveal a strong positive correlation between remittances and economic indicators, such as GDP growth and foreign exchange reserves, underscoring their role in poverty reduction, women's empowerment, and employment generation.

9. Statistical Analysis of Demand for DTH Services among TV Users in Narayanganj City through Direct Survey Method 

   • Supervisor: Mohammad Saif Noman Khan, Assistant Professor, IBA, University of Dhaka. (Alumni: Schulich School of Business, Canada) 

   • Collected 500+ field data and conducted an in-depth analysis of survey data employing SPSS to model data distribution, perform Customer analysis, and analyze Product Design aspects. 

10. User Feedback on Ride-Sharing Services in Dhaka to Understand Customer Preferences through SPSS Data Analysis 

    • Supervisor: Avijit Mallik, Assistant Professor, IBA, University of Dhaka. Fulbright Alumni (University of Texas at Austin) 

    • Developed an analysis workflow comprising regression analysis, one-sample and two-sample T-test, Chi-square test, etc. that determines customer preference regarding ride-sharing services. 

11. Data-Driven Exploration of the Effects of Online Reviews on Purchasing Decisions among Young Adults in Dhaka 

    • Supervisor: Dr. Rezwanul Huque Khan, Professor, IBA, University of Dhaka; PhD (University of Warwick, UK) 

    • Performed research through an online survey questionnaire and applied descriptive and inferential statistical tools with R, SPSS, and MS Excel to determine what platform young adults of Dhaka check for online reviews, how much they trust each platform, and for what products they search online reviews for.