North South University-এর শিক্ষার্থীদের প্রজেক্ট, লেখা, গবেষণা ও কার্যক্রম।
প্রজেক্ট
Here I created two interactive dashboards to gain valuable insights into the transactions and customer behaviour with card payments. Analyzed the data to monitor key performance metrics to understand the trends.
প্রজেক্ট
Used Pandas to explore data and understand the demographics, also used Matplotlib to plot better and visualize the data better.
প্রজেক্ট
In this SQL project I explored the data to analyze and better understand the situation of layoffs in companies. Used queries to effectively understand and understand the demographics.
প্রজেক্ট
In this project I cleaned the data in SQL. In this case I used queries to clean and analyze data as I explored the data.
প্রজেক্ট
In this project I analyzed the customer behaviour by first using python to clean and standardize the data. Then using SQL queries, I tried to get an idea of the customer behaviour using queries, and then used Power BI to visualize the data and created das
প্রজেক্ট
An automated electricity billing management system designed to calculate electricity usage and generate organized customer billing records, demonstrating automation, information management, and problem-solving.
প্রজেক্ট
A digital information management concept designed to help organize and manage patient-related information electronically, with emphasis on structured data organization and user-oriented system design.
প্রজেক্ট
The Future of Voting and Elections: 'Amar Vote-An Online Voting App' (under publication of North South University Research Centre )
প্রজেক্ট
Compared LSTM, Bi-LSTM with custom Attention, and Transformer architectures for stock price prediction; deployed as a live dashboard with real-time Yahoo Finance data. Bi-LSTM + Attention achieved best performance: RMSE = 4.13, R² = 0.9556 on out-of-sampl
প্রজেক্ট
Built an end-to-end pipeline on 5,351 patient records using pre-operative features only, with strict leakage prevention and SMOTE to handle 2.7% class imbalance. 97.5% accuracy, AUC = 0.813 — used SHAP for clinical explainability; deployed the final model
প্রজেক্ট
14-disease chest X-ray detector: trained DenseNet121 on 89,696 NIH ChestX-ray14 images, reaching mean AUC 0.7765; Grad-CAM heatmaps visualize model focus regions. Deployment: LLaMA (via Groq API) auto-generates structured radiology reports; full-stack dep
প্রজেক্ট
A drive aggregator that pools total storage of google drive and dropbox
প্রজেক্ট
Developed an e-commerce platform including product listing, shopping cart management, checkout, user authentication, payment gateways, and order processing functionalities. Tools used as Frontend: HTML, CSS Tools used as Backend: Python, JavaScript
প্রজেক্ট
Designed to enhance task organization, collaboration, and productivity within a team. Tools used as Frontend: XML Tools used as Backend: JAVA, Firebase, Google Fit API
প্রজেক্ট
Developed an alarm app with onboarding, location access, and alarm scheduling. Implemented scheduled local notifications and local data storage.
প্রজেক্ট
Evaluated classification across seven FairFace groups using per-class accuracy and confusion matrices.
প্রজেক্ট
Built an SVM pipeline to classify parking spaces in video and overlay available-space counts. Used mask-based connected components to identify spaces and calculate the nearest free spot.