An individual project for implementing a solution for Pump it Up: Data Mining the Water Table competition in DrivenData. Followed pre-processing, encoding, feature engineering, and used models like the Random Forest, Decision Tree, Gradient Boosting Classifier, XGBoost, and MLP Classifier

Used models

• MLP Classifier
• Random Forest
• Decision Tree
• XgBoost
• Gradient Boosting Classifier

An automated Radiography analysis framework for Pneumonia and Covid-19 identification can be used to provide better performance in chest x-ray analysis for detecting lung infection conditions.

A systematic review presenting deep learning-based pneumonia and coronavirus detection solutions, trends, datasets, guidance for a deep learning process, challenges and future research directions.

Chest x-ray classification with MobileNetV2, InceptionV3, Xception, ResNet50 models with added top layers

Develop an ensemble using the earlier trained MobileNetV2, InceptionV3, Xception, and ResNet50 models

experiment chest x-ray classification with segmentation using U-net architecture

Develop a web application for the developed framework consisting of segmentation and classification as a proof of concept.

Publications

Journals

D. Meedeniya, H. Kumarasinghe, S. Kolonne, C. Fernando, I. Díez and G. Marques, ”Chest X-ray analysis empowered with deep learning: A systematic review”, Applied Soft Computing, p. 109319, 2022. , DOI: https://doi.org/10.1016/j.asoc.2022.109319

K. A. S. H. Kumarasinghe, S. L. Kolonne, K. C. M. Fernando, D. Meedeniya, ”U-Net Based Chest X-ray Segmentation with Ensemble Classification for Covid-19 and Pneumonia”, International Journal of Online and Biomedical Engineering (iJOE), Vol. 18, No. 7, pp. 161-174, 2022. DOI: https://doi.org/10.3991/ijoe.v18i07.30807

Conferences

C. Fernando, S. Kolonne, H. Kumarasinghe and D. Meedeniya, ”Chest Radiographs Classification Using Multi-model Deep Learning: A Comparative Study,” 2022 2nd International Conference on Advanced Research in Computing (ICARC), 2022, pp. 165-170, DOI: https://doi.org/10.1109/ICARC54489.2022.9753811

S. Kolonne, C. Fernando, H. Kumarasinghe and D. Meedeniya, ”MobileNetV2 Based Chest X-Rays Classification,” 2021 International Conference on Decision Aid Sciences and Application (DASA), 2021, pp. 57-61, DOI: https://doi.org/10.1109/DASA53625.2021.9682248

This project is a distributed chat application. The project's goal is to create chat servers where clients can connect and chat with one another. Chat servers offer clients to connect with unique ids, create/join/delete chat rooms, chatting with other clients, etc. For leader election, we have used bully algorithm, for failure detection we have used heartbeat. We have used the Netty framework for TCP connection.

My Role

I developed client requests for deleting room, room list, moving from room and joining for another room. Also done part of the bully algorithm.

A COVID management tool that has authentication, bed reservation, COVID statistic review. This applications' REST API was built using JAVA Servlet, the UI was built using React with React-redux, and PostgreSQL was used as the database. A web application for COVID management in Wakanda. Additionally for the authentication system was developed using JSON Web Token (JWT), and for password hashing used jbcrypt. For logging functionality, Log4j has been used.

In this project, a web-based system and two mobile applications, with REST API as the backend were developed where the passengers can reserve bus seats online by paying the amount. This will facilitate passengers and conductors, making the ticketing process more convenient. A mobile application will be developed for the driver/conductor to handle the ticketing process from their side. The web application will facilitate both bus owners and passengers. After paying the relevant amount through the web app or the mobile app the passenger should receive a receipt confirming the reservation and the conductor/driver should receive the information about the reservations.

My Role

I developed the REST API by using Node js with Express and designed the database(firebase) of the system. This was the first REST API I built. I gathered a lot of experience from this project and was able to use new tools like Postman, Apache JMeter.

A web application for school administrators and a mobile application for school teachers, intended to improve the efficiency of the school system. This system would help to monitor the attendance of both teachers and the students, teachers could take their notes, teachers would receive the notifications reminding the next period.

My Role

I developed the mobile application of teachers using flutter.This was a minimum viable project build for Code with WIE competition and this was my first mobile application. By developing this mobile application I was able to get experience in mobile application development.

Achievements