This project is completely up to you!* *terms and conditions may apply. Consult your TA or lead teacher for full details of this limited offer
- Project Description
- Project Goals
- Requirements
- Deliverables
- Mentoring
- Schedule
- Presentation
- Tips & Tricks
- Resources
In this project, you will think of a topic and problem, collect experimental data, complete an end-to-end analysis and present the results, all by yourself.
First, choose a topic of interest to you and understand what research has already been done in that area. What are some interesting questions that remain? Can you turn those questions into a product (i.e. can you extract value out of answering those questions)?
You will then collect some data you think could help answer those questions. Choose your main source of data wisely, since in this project you have a restriction that tries to emulate a common corporate setting: you won't have access to a census of the universe of your choice. You must collect the data yourself in such a way that the universe of datapoints available to you is limted. For example, you may be limited by time (e.g. watching and categorizing Youtube videos or Instagram pictures), by cost (e.g. querying Google Maps for public transport routes via the gui, without paying for the API access) or by access (e.g. surveying people on their preferences). In the end, you should aim at collecting between 30 and 100 observations (rows) and between 5 and 10 features (columns) per observation.
Once you have your data, complete an analysis that answers your original question and/or related ancilliary questions. Please make sure that the main observations you make hold to scientific scrutiny at some level of significance. You can and should supplement your analysis with visual intuition and highlights of hypothesis that the data seems to support, even if you are not necessarily able to hold those insights to the same level of scrutiny as your main question.
You can enrich your limited dataset with information from richer sources that you can obtain trough any means you've learned before (e.g. you may web scrape the weights of car models if that is one of your observations).
Like in the previous project, package your results with a product or service mindset. You will present your findings in a presentation (possibly supported by an interactive visualization) where you should evidence principles of dashboarding and storytelling.
- Research, collect and analyse data on a topic of interest to you.
- Feel free to use additional data to enrich your dataset, maybe using an API or web scraping.
- Apply the statistical techniques we have learned, along with techniques from EDA.
- Create useful and easily-interpretable plots.
- Prepare a presentation keeping in mind the finer points of storytelling.
- Communicate the results of your analysis clearly, accurately and engagingly.
- You must plan your project. That is why creating a Kanban or Trello Board is mandatory. You have a template for Trello here.
- You CAN'T CODE until you project is planned.
- Create a .gitignore file and include it in your repository.
- All the scripts you used for your analysis.
- Slides and a 5 minute presentation in the classroom.
- Repository with your workflow + documentation + code. Even if you are working alone, you need to maintain good practices!
- A short report including your motivation, methodology and results.
One of the TAs will be your mentor! Your mentor will:
- Follow your project in general.
- Check if you are following the tasks, your blockers, etc
- Help/support you in specific questions.
Monday
- Think about a topic and propose some core questions.
- Choose data that is relevant to your questions and devise ways of collecting such data.
- Choose ancilliary data that would allow you to acheive your strech goals.
- Look for documentation to give context to your project.
- Write the README file in your repository.
- Get approval for your project
- DO NOT START CODING
- Start collecting the data for your core questions
NO CODE UNTIL HERE
- Tuesday - Thursday morning*
- Data entry, cleaning and transformation.
- Start the analysis. Remember all the techniques you have learned!
- Prepare a draft of your first slides presentation (no analysis or conclusions yet): title, motivation, context, ...
Thursday afternoon
- Rehearsal. Take the feedback and use it!
- Finish the analysis. Finish the slides.
- Final improvements!
Friday
- Presentation!
Presentations for this project will be in the classroom! Presentations will be EXACTLY 5 minutes long, with 2 additional minutes for questions. We will stop you!
- Organize yourself (don't get lost!).
- Ask for help vs Google is your friend.
- Define a simple approach first. You never know how the data can betray you ;)
- Learn about your subject and understand what other research has been done before you.
- You can use data from the projects your partners did in the last weeks.
- Before making a graph, think about what you want to represent.
Here are some data sources that could be interesting to you: