Assignments Due

Week 2 checkpoints should be submitted by Monday, 01/15/2018 at 11:59 PM PST.

Perform

Group 1: Capstone Critique (50%)

What are the ingredients to data science quality? It is often easier to answer this question through a critique of existing work than to come up with a list off the top of our heads. This assignment asks you to look at MIDS Capstone Projects from the past several semesters to learn from the efforts of recent MIDS graduates. You will apply your own critical eye toward what you think works and does not work about their data science concepts and their presentation styles.

First, read all of the instructions for this week and next week. Connect with your team and schedule a meeting to review the instructions together. You can meet in your ISVC room or on Slack (which does have video and leaves a more accessible chat record).

As a team, choose two projects that you find interesting or promising from the lists presented on the MIDS Capstone Projects pages. Consider having each team member nominate their two favorites and discuss the short list together.

Starting with the slide deck template, provide a side-by-side comparison of each project.

  • (1 slide) Always have a title slide! How will you pique our interest?
  • (1 slide) An introduction to and basic background about each project.
  • (1 or 2 slides) Decision Making Criteria:
    • Compare the two projects according to the list of criteria in the Decision Making section of the DSM Template.
    • Decide which project fared better on the criterion and why (e.g. who defined the problem more clearly?). You don’t have to cover all criteria, and you shouldn’t do more than two criteria per slide.
  • (1 or 2 slides) Organizational Design Criteria:
    • Follow the same process using items from the Organizational Design section of the DSM Template.
    • Again, who did better and why? No more than two criteria per slide.
  • (1 slide) Describe how either project did something really well (or poorly) that is not asked about on the DSM Template. Label the “missing” criterion, file it under the appropriate heading (e.g. Decision Making, or coin an entirely new category), and form a question that would help us to identify it.
  • (Q&A) After you present, open discussion by telling the audience which project your team thinks is more meritorious and asking whether they agree. If one had to win it, who deserved the Hal R. Varian MIDS Capstone Award?

Your source material for assessing each project must include the video recording, if available, of the original teams’ presentations as delivered in the semesterly Capstone Showcase. Otherwise please use all available documentation provided through a team’s project page. If you are feeling ambitious and the original team left an archive or repository of an executable app or script that you can deploy, go ahead and test it out, but don’t waste time debugging it if it doesn’t work.

How to submit

One member of the team should make a copy of the slide deck template and upload it to the submission folder. The team should collaborate on that document for this week’s checkpoint and next week’s final submission.

Come to live session having already met once as a team and with a rough draft of at least your background, decision making, and organizational design slides. Keep track of any questions that come up and bring them to class.

Additionally, select one member of the team to host the presentation recording in her Adobe Connect room. The host should create a test recording, using a PDF of the slide deck template, and record a 15 second test video similar to the one made in the pre-recording tutorial. Upload the test video to this bDrive folder by this week’s deadline. Keep track of any issues you experience and we will help debug in class.

Compose

Individual 1: Big Idea 1 (25%)

How should you develop and explain a data science idea? It is as simple, and as difficult, as writing it down. Give your idea a beginning, a middle, and an end. Introduce a problem and whose problem it is. Propose a solution. Describe the difference it makes. A beginning, a middle, and an end. A problem, a solution, and a difference. It’s a structure that is easy to learn but hard to master. So sit down and write, and try to pour your creativity into that simple mold.

Communicating your data science ideas starts with some basic building blocks. First, you need to have a handle for your idea, some answer to the cocktail party question, So what have you been working on? This handle appears in the title of the write up of your idea, and in a slug which is the briefest label you can think of as a badge for your concept. Let these two elements be catchy, and more importantly, let them already be informative. The name for your idea is the real beginning before the actual beginning, since it is an audience’s first glimpse of what is coming.

Once you can introduce your idea then you need to begin to explain it. The standard convention and our formula in RDADA is to start with a description of a worthwhile problem and one that is facing a particular kind of person or organization. This is the beginning.

Once we understand the problem and the urgency behind it, it is then time to describe the solution. Here is where you can unleash your data science imagination. Suggest an analytics pipeline that can turn noise into signal, information into knowledge, or raw data into a product or service. Be ambitious and do not be afraid of what you have not yet learned. It is enough to name methods and describe what they promise to accomplish. You do not need to explain how they work. This is the middle and should be the longest part of your composition.

After your best attempt at a solution, the work is not over. All solutions are promises or proposals, and you must make good on these by describing the consequences, intended and unintended, of implementing your solution. Trumpet the accomplishments, even if only as expectations, but also be responsible for addressing the limitations. No solution will be a slam-dunk, and you need to be prepared for exposure to unintended consequences. Be measured in your discussion of outstanding issues and always end on a positive. This is your end.

GitHub Classroom

Exposition isn’t a mystery, but the work remains to be done. Your task for the first individual assignment is to lay out the above components in a development environment that will make revision, publishing, and collaboration easier. We will be using GitHub and a variation of Markdown to model a productive workflow for writing in plain text. I’ve prepared a template for a bookdown project, which you need to add to your GitHub account by following the link below:

https://classroom.github.com/assignment-invitations/c5b9454ac138cd70f5ddeb2cd028028f

After you follow the authorization instructions it will fork the repository automatically, and you should see a new repo under your GitHub account and under the organization w201rdada. Your repo will have the address https://github.com/w201rdada/portfolio-yourgithubusername/, though you may have to wait a minute for it to propogate. This repo is private and will not count against your private repo allotment. At the start only you and myself as the organization owner will be able to see it. Eventually you will need to add others in the class as read-only collaborators.

How to submit

Your submission will be due as a commit to a GitHub Classroom repository that will continue to serve as your student portfolio for the rest of the term. Consult the basic submission workflow tutorial to learn how to make a commit to your GitHub.com remote repository.

When you initialize your repository you can view the files in it through GitHub.com. The only file that you should concern yourself with for Big Idea 1 is the one called 01.Rmd. We will explore the rest later. You don’t even have to clone the repository locally to submit your work, but you should at least copy and paste the contents of 01.Rmd into a plain text file on your computer.

To complete the first checkpoint, start replacing the boilerplate with your own writing. Concentrate on writing a title, a slug, and headings for the beginning, middle, and end. These elements should serve as a high level outline for your idea. Take one extra step and write one sentence to replace the hipster ipsum in each section. That’s one sentence for your beginning, one for your middle, and one for your end. You have to start somewhere! If you’re inspired, by all means keep writing, but the minimum viable product here is a three sentence essay with a title and headings.

By next week’s checkpoint you’ll need a first draft in the ballpark of 300-500 words, but for now just get the ball rolling with your initial ideas.

Review

No checkpoint due this week :)


readings (links) & lectures ～ assignments due