Skip to content

Mini projects

Jump to bottom
JonnaBuri edited this page Oct 26, 2017 · 14 revisions

Mini-project Instructions

26/10 UPDATE - GROUP PITCH SESSION LOCATIONS [see right below]

There will be 3 sessions running at the same time, in three rooms. You should show up at your coach's session:

  • Chang: Exactum, CK112
  • Ville-Veikko: Exactum, D122
  • Johanna: Exactum, D123

Overall Guidelines:

  • To get started, follow the guidelines for week 5 [see below].

  • 10/10: Submit [as a group] the "Create a topic" assignment on peergrade by 10/10.

  • After you get the feedback from your coaches on your topic, you can proceed with the development of your project.

  • Set your milestones as a group and get to work!

  • 26/10: The mini-project presentations will be held on 26/10, at 16.00 [date and time that the exam would originally take place]. There will be 3 presentation groups [one for each coach group] and they will be held simultaneously in 3 rooms. Don't worry, you will hear more info about this.

  • 27/10: Is the last day for you to hand in [as a group] the "Project" assignment on peergrade. And peer review begins. Each student should review 2 submitted projects, based on the criteria you will find on the peergrade review form.

  • 31/10: Is the last day for you to review your peers.

Week 5 - Getting ready for your mini-project!

  1. Form groups of 3: There is a large variety of people available: from Mathematicians to Computer Scientists [and others, I'm sure!], make sure to mix and match, as to form a powerful group of 3! You don't have to be in the same exercise group! *In case no matches are available, we got you covered! Just let us know at the exercise groups!

  2. Pick your coach: Every group should pick their coach. Again, it doesn't have to be the one from your exercise group, but this is the person you will be able to turn to for questions & guidance and they will be monitoring the progress of all the mini-projects in their group. So you can go here: https://doodle.com/poll/9r2thruuvhq9pnyq and register your group to a coach with available positions. One person per group votes and represents the whole group. Take any pick from: Chang, Ville, Johanna!

  3. Register to peergrade: Every student should go and register as a student here: https://peergrade.io/join using the class code GBCDW7. After you have registered as a student, go on and form your group on peergrade!

  4. Coming up with the topic: Brainstorm and research on the topic of your mini-project. The outline of your project description should be as follows:

             A. Title of the project	
         
         B. Elevator pitch [max 400 characters]

         Describe the data science challenges related to the following themes. 
         A good project should include interesting challenges under most of the themes.
         So for example, a project that begins with a single clean data set and a clear task 
         (e.g., predict a given response), isn't a good project.		
         
         C. Data: sources, wrangling, management		
         
         D. Data analysis: statistics, machine learning	
         
         E. Communication of results: summarization & visualization
         
         F. Operationalization: creating added value, end-user point of view

This is actually the most challenging part: coming up with the problems yourselves. But worry not, we got your back!

You will find further below, some example mini-project topics, to get an idea of what kind of a problem you should come up with!

  1. Submit your topic idea [deadline: 10/10]: This week's assignment on the mini-projects is to have your topic submitted as a group on peergrade by 10/10, under the assignment "Group Project - Create a topic", so that the coaches can go over your proposals and give you some feedback on the outlined topics. You will get the feedback on your outline within a couple of days, and then you'll be able to proceed with the development of your project!

Instructions for Solo Project - exam

We very much recommend the group projects. If the timetable is absolutely impossible, you can also choose to take the alternative option which is to submit an individual project (so a "Solo Project") and take a separate exam.

Concerning the Solo Project, pretty much the same guidelines apply for you, except:

  • Your deadlines for submitting the topic outline are the same with the group projects, but use the assignment "Solo Project - Create a topic".

  • Please do not register to a coach via Doodle, or there won't be enough room for the groups to register there. Contact any coach and they can help you with any questions, unless they are completely packed with projects, in which case they will ask you to turn to another coach.

  • THE PITCH SESSION FOR SOLOS will take place on November 1, 2017 12:00-14:00, room D122 in the Exactum building!

  • REMEMBER TO REGISTER to the separate exam. The first separate exam is November 29, 2017. You must register at least ten days in advance to be able to take the exam. No late registration is allowed. (A university level rule.)

Assessment Criteria

Some key-points to be taken into consideration for your work. [You will also review your peers' work according to these criteria, so pay close attention and keep notes at their pitch too!]

  • Presentation and pitch: This is the overall impression you get from the presentation of the problem and its solution. How clear is the problem description? Are the results communicated well? How convincing was the pitch? [depending on the topic's purpose]
  • Value: What is the value of the final deliverable? How useful are the results of the project?
  • Effort: Do you think there was a sufficient amount of work put into the project?
  • Idea: How original is the topic idea in your opinion? How useful/innovative is it? Could it have more applications than the one described in this case?
  • Topic scope: How suitable was this project for this course? Are all stages of the data life cycle taken into account?
  • Actionable: How actionable is the outcome? Could it be used in real life applications?
  • Deliverables: How understandable is it to your target group? How understandable is it to everyone? Could it stand alone without your pitch? Needs to describe what you did and how you did it in a non-technical manner.
  • Project analysis: Data wrangling, data analysis, communication of results will be taken into account, based on each project's topic. For example, the visualizations of a project: How well do the visualizations communicate the point made? Could correlations between other variables have clearer results etc

TA Contact Info

Clone this wiki locally