Projects
Near Earth Objects Danger Risk Prediction
This project aims to predict whether a Near-Earth Object (NEO) is potentially hazardous using machine learning. Leveraging a dataset of NEO observations from 1910 to 2024, it classifies objects based on various features into hazardous or non-hazardous categories, providing a robust solution to enhance NEO risk assessment.
Python • Pandas • Scikit-learn • PyCaret • Streamlit • Lime • Skops
Predicting Dengue Fever Outbreaks in Sri Lanka
As part of an Omdena project, I contributed to developing an AI-driven system to predict dengue fever outbreaks in Sri Lanka. By applying time series analysis and machine learning techniques, the project aimed to enhance early warning capabilities, enabling more efficient public health responses and potentially saving lives.
Python • Pandas • Scikit-learn • Darts • Streamlit
ConnectCam
ConnectCam is a full stack web application that is built for the purpose of video conferencing. It is a simple and easy to use application that allows users to create a room and share the room link with others to join the room. It is built using WebRTC and Socket.io for real-time communication.
Next.js • Node.js • Express.js • WebRTC • Socket.io • TailwindCSS
MediLeaf
MediLeaf is an application whose motive is to help the individual to identify medicinal plant with their properties by just scanning the leaf of any plant which might result creating curiosity about plant that lead to the preservation of the valuable plants as well as source of income.
Flutter • Tensorflow • Django • PostgreSQL
Freyja
This is a web app to grade essay using LSTM model using Django REST, React and Tensorflow. The dataset used to train the NLP model is https://www.kaggle.com/mpwolke/hewlett-foundation-essay-scoring , this project was the education category winner of Quantum Hack 2020.
Django • Python • Tensorflow • Spacy • NLP • Docker