Hacking Unemployment - Data Science Challenge 2022

A Series of Global Data Science Challenges

The Fall 2022 IronHacks Training Challenge

Learn how to to use Jupyter Notebooks and BigQuery for exploring new data science techniques and models using Indiana Department of Workforce Development data to predict unemployment claims in Indiana!

The IronHacks challenge runs for three weeks, so it is geared towards learning and iteration. This means that even if you are new to data science it is an excellent opportunity for you to learn by doing! Further, this challenge focuses on an important topic: predicting unemployment trends.

The task: Predicting unemployment claims in the State of Indiana using Department of Workforce Development (DWD) data

We ask you to predict the number of unemployment claims using historical data provided by Department of Workforce Development (DWD), as well as demographics data. You are given variables such as race, education, categories of employers, week number, etc. Your task is to predict the total unemployment claims for a specific census tract in Indiana in a specific week. Using such data, your predictions will inform citizens as well as decision-makers like the State of Indiana, the Department of Workforce Development, and others about the potential impact of unemployment to the state.

Your workspace

During the competition, you will use the new IronHacks workspace that supports a fully integrated data science workflow: You will have access to a JupyterLab and libraries to query and analyze data from BigQuery. Please note: For this competition, you will only be using Python. Please check out the workspace by clicking on the workspace button in the right corner.

Accessing the data in BigQuery

Our Tutorial #2, which will be available when registration at ironhacks.com opens, will tell you more about how to query the data.

Your submission

You will be asked to create a predictive statistical model that you can revise over time. You will be asked to produce and submit your results in a Jupyter notebook. You can learn about the details of your submission in the task statement when it is released to all participants. You will have multiple opportunities for making a great submission. The task statement will detail the submission frequency and requirements, but generally, we strongly recommend you to submit often.

Evaluation

Each participant’s notebook will be evaluated in two dimensions using objective metrics:

  • Prediction accuracy (your model’s MSPE and MAPE)
  • Individual exploration and learning (your effort to explore a variety of different models)

The evaluation metrics and logic will be detailed in the task statement once it has been published. Please also carefully read the rules at ironhacks.com when the competition is open for registration at ironhacks.com.

The process

  • Registration and warm-up phase: When you register for the hack, you will have the opportunity to warm-up using our tutorials. You will also access to a subset of the data that you can use to get familiar with the data you will be using in the challenge. See the table below or the calendar on the IronHacks website when you register at ironhacks.com.

  • Your start of the Competition: The competition starts with you accepting the rules at ironhacks.com of the competition. Once we have released the task, you can start querying BigQuery.

  • Multi-staged interim submission process: After the release of the task, there are multiple submission opportunities over a three weeks period of time. There are three submissions planned for each week (one mandatory submission and two submission opportunities) so you can be scored three times for the same model focused on a particular week. We will get new data every week so you have the opportunity to tune your model every week further, as you get new data.

We will create a Calendar for you to learn more about the dates and deadlines below.

EventDate/Time (EST)
Registration opensOctober 21, 2022 6:00 pm
Task is releasedOctober 22, 2022 6:00 pm
First Data ReleaseNov 2, 2022 11:59 pm
First submission opportunityNov 3, 2022 11:59 pm
Submission opportunitiesEvery Thursday and every Saturday at 11:59 pm
Mandatory submissionsEvery Monday at 11:59 pm
Update of database and updated target week for predictionEvery Tuesday at 12 pm
Registration EndsNov 21, 2022 11:59 pm
Last Mandatory SubmissionDec 5, 2022 11:59 pm

Skills and tutorials

This IronHacks is designed for participants with intermediate knowledge of statistics and Python. Beginners and more advanced users are also invited to participate. If you have never worked with Jupyter notebook, you can use the warm-up phase to practice. You will have access to check out the tutorials once registration at ironhacks.com is open. We will release more tutorials as needed.

Prizes and awards

For the Fall 2022 hacks, we award several prizes with a total value up to USD 2000. More details will be available in the upcoming Task Statement.

In addition, there are more additional benefits for you:

  • Certificate: Every participant will receive a digital certificate for participation with logos of the sponsors.
  • Showcases: We will publish selected models www.ironhacks.com at the end of the competition. In addition, our partners will also feature the best solutions on their website.
  • Learning experience: Based on what we have learned from the past, IronHacks will offer a great learning experience and practice for data scientists.

And do not forget - your work is essential to help us contribute to knowledge and solutions surrounding unemployment.

Questions?

After registration at ironhacks.com, just check out or leave a note in our forum! There you find announcements. You can also contact us by clicking on the “Get in touch” button in the upper right corner.

Frequently Asked Questions (FAQ)

  • Where can I find the tutorial material if I missed a live tutorial? You can find the recordings as well as our demo notebooks once you register at ironhacks.com.
  • Where can I find recordings if I missed an info session? Recordings of our weekly live sessions/info session will be posted in the forum as links to our youtube channel here. So we suggest you subscribe to this channel to have updated info about that!
  • How do I get access to the competition? Hack the Unemployment Challenge requires you to consent to the rules of the competition here at ironhacks.com. Once you have provided consent to the scientific study accompanying this challenge you are eligible to participate.
  • How do I know when the deadline is for each submission? Check out our calendar to learn more about the submission deadlines or post in the forum.
  • How do we access the data? You have to follow the instructions in tutorial #2 to access the data in a notebook.
  • How do I contact the IronHacks Team? You can post in the forum or send an email to us at [email protected]
  • Can I create a team? No, hacking unemploynment is an individual challenge.