Congratulations! You have been hired as a data scientist at Centers for Disease Control and Prevention, United States of America. You will be part of the newly established data science department. The government has collected data on covid related cases, deaths starting from January 2020. Your supervisor would like to better understand how the data can inform good public policy for the next pandemic. You and your team will put your data analysis skills to the test by uncovering insights about the pandemic.
However, your supervisor isn't a scientist and doesn't fully understand what kinds of insights this dataset can offer. So your team has two tasks:
First, your team should define a set of question(s) to explore with your data. These are well defined questions that the data can provide an answer to. Second, your team should provide an answer to this question. The answers will depend on the question(s) and could be visualizations, models, databases, etc. Teams will be judged both on their questions (scope, impact, relevance, originality, etc) and answers (technical skills, correctness, quality, usefulenes, etc.) Here are some possible ideas to use for inspiration...
Trend Analysis:
- How have COVID-19 infection rates changed over time in different states or counties?
- Can you identify any correlations between policy implementations (like mask mandates or lockdowns) and changes in infection rates?
Predictive Modeling:
- Can you build a model to predict future COVID-19 case trends based on current data?
- What factors are most predictive of COVID-19 outbreaks in a region?
Impact on Public Health:
- What has been the impact of COVID-19 on hospital capacity and resources in various regions?
- How does the spread of COVID-19 correlate with other health outcomes like mental health issues or delayed treatments for non-COVID conditions?
Socioeconomic Effects:
- Analyze the socioeconomic impact of COVID-19 on different communities based on income, race, or employment sectors.
- How have educational outcomes been affected by COVID-19, particularly in relation to remote learning adoption?
Geospatial Analysis:
- Map the spread of COVID-19 across different regions and identify hotspots over time.
- Can you find geographic factors that influence the spread or containment of the virus?
Comparative Analysis:
- Compare the effectiveness of different health policies in managing COVID-19 spread across states or countries.
- What lessons can be drawn from regions that successfully managed their COVID-19 response?
Vaccination Analysis:
- What is the impact of vaccination rates on COVID-19 case numbers and mortality rates?
- How have vaccine rollouts affected the dynamics of the pandemic in various demographic groups?
Sentiment Analysis:
- Use social media data to analyze public sentiment about COVID-19 vaccines or government policies.
- How does public sentiment correlate with actual COVID-19 case trends?
Machine Learning Applications:
- Employ machine learning techniques to classify areas as high or low risk for future outbreaks.
- Can you develop a system to recommend interventions based on local COVID-19 forecasts?
Healthcare Resource Allocation:
- Model optimal distribution strategies for medical resources (like ICU beds or ventilators) based on COVID-19 case forecasts.
- What strategies can be implemented to improve healthcare response in underserved areas?
These prompts are designed to provoke thought and creativity, allowing teams to explore a wide range of data science techniques and applications. We strongly encourage teams to formulate their own unique questions based on their interests, skills, and what can reasonably accomplished in 4 hours.
- Dataset : New York Times Data on Github
- Getting started jupyter notebook
- Github page
- Organizers :
- Download the dataset and process it mentioning any problematic entries.
- Build and validate your models.
- Use your model to answer questions.
- Explain how you validate your model
- Bonus task : use the county level data to predict a regional lockdown