Datathon'20 - Runner up

Jul 28, 2023   | #Machine Learning  #Temporal Modeling  

Competition Title	Datathon 2020, All island Competition
Organizer	SLIIT Campus, Sri Lanka
Year	April 2020
Place	2st place
Team	Devin, Tharindu, Nipun
Code	Paper

Description

Our participation in the All-Island Data Science and Machine Learning Competition was a journey of data exploration, innovation, and societal impact. In this thrilling competition, we embarked on a comprehensive data analysis quest that led to a groundbreaking solution. Our final challenge was to design a system for detecting Covid-19 spread and conducting a thorough analysis of its impact. We took a multifaceted approach, crafting an LSTM-based multivariate time series model for precise Covid-19 spread predictions at the district level. To provide valuable insights, we harnessed meta-heuristic analysis techniques to evaluate the potential impact of Covid-19 on various selected industries.

Technologies

Methodologies

LSTM Modeling Auto Regression Modeling Automated Pipeline deployment

Key Achievements

• We created a ETL pipeline to web-scrape and retrieve covid-19 statistics to feed into our models
• We successfully developed a sophisticated LSTM-based multivariate time series analysis model, showcasing our prowess in this cutting-edge technology.
• Our Covid-19 spread detection and impact analysis system aimed at making a positive difference in society, emphasizing the importance of data-driven decision-making.