Reviewing data science related MOOCs

Taking stock of the MOOCs I tackled while on paternity leave.

Thanks to a generous (for Switzerland) paternity policy, I had 10 weeks home with my son Timo. He though decided to spend a good chunk of that sleeping, which meant I had plenty of time to update my site, and spin up two new sites (an internal team site, and I refactored rinpharma.com to have a new theme).

I also spent a lot of his naps working through 11 Coursera courses on my todo list. This post is recapping which of those courses I recommend - as I’ve already been pushing a few to my colleagues as great courses they should do.

As a tldr - if you are a data scientist using AWS - I can highly recommend DevOps Culture and Mindset and AWS Cloud Practitioner Essentials as 2 courses you should consider.

Table of Contents

Ways of working

DevOps Culture and Mindset

  • Link and my Certificate
  • UC Davis (Coursera MOOC)
  • I feel like this course did a great job making me think about things I’d come across, but hadn’t taken a step back to think through as important points to structure how we work. All data scientists should be aware of concepts like decoupled architecture, CICD, feedback loops, and have opinions on how to structure teams and projects to manage work and risk. A definite 👍 from me and I would view this as one of the courses I’ve taken that made me think about the way my team worked, and could evolve, the most.

Agile with Atlassian Jira

  • Link and my Certificate
  • Atlassian (Coursera MOOC)
  • We used Jira in my team as a way to track research, which is using Jira in a different way than how it was designed. I learnt from someone that joined our company from a tech background that we did a pretty bad job at designing our Jira project flow - so was keen to understand the tool better and took this course. Unfortunately, I didn’t feel like I gained much from this course. If you are planning on running sprints on day 1 of a new job and your old company used a different tool, it would be worth taking this. But the course content is quite simple and despite hoping to learn some advanced and improved ways of using the tool, I felt I picked up more from prior experience working in a team that uses Jira and being involved in a few ‘agile’ sprint based projects.

Managing data science teams

Building a Data Science Team

  • Link and my Certificate
  • Johns Hopkins (Coursera MOOC)
  • A really quick course (~1.5 hours of videos). It talks about skill diversity in hiring, onboarding, structure (e.g. centres of excellence vs embedding), stakeholders and I guess most importantly your purpose of enabling users to gain insight from data. I personally disagreed occasionally with some of the content as it was overly simplified. An example being breaking down the domains into only data science manager, data scientist and data engineer. I much prefer to think of it as a very blurred line (e.g. data engineering is also the responsibility of data scientists and informatics/IT). I guess I’d recommend to take this course, but I preferred the dev ops course for any topic related to team structure and workflows.

How to manage a remote team

  • Link and my Certificate
  • Gitlab (Coursera MOOC)
  • I did this course, then the devops course. So I loved it, then realised there was greener grass out there. Would recommend this one if you were actively looking to pitch for a remote team (writing that pitch is what the whole course centers around…), but as I managed people based in the UK and EU before this course, I didn’t gain too much beyond what I already had from the discussions and support I received from my peers and company.

Tech

AWS Cloud Practitioner Essentials

  • Link and my Certificate
  • Amazon (Coursera MOOC)
  • This course was really fantastic. It’s like an entertaining walkthrough of the AWS catalogue, and if your data science team use AWS - I’d almost say this is a requirement so you can have constructive conversations with your informatics partners. Note that it doesn’t give you any details - so this gives a data science a grounding to understand the general AWS landscape. It is in no-way giving you the information you need to architect your analytic platform on AWS.

Analytics

Design and Interpretation of Clinical Trials

  • Link and my Certificate
  • Johns Hopkins (Coursera MOOC)
  • I’m an epidemiologist, that worked on a Phase IV trial for my PhD, and in a team at Roche on the periphery of trials doing real world evidence for 5 years. So I took this course thinking it would ground me a bit more in the clinical trial world. Starting this course I quickly realised it was taught by epidemiologists, who seemed more removed from Phase I to III in pharma then myself. There is a use to this course, but definitely only to those with no background at all in the area. The course is along the lines of clinical trials 101.

A Crash Course in Data Science

  • Link and my Certificate
  • Johns Hopkins (Coursera MOOC)
  • The title basically says it all.

Managing Data Analysis

  • Link and my Certificate
  • Johns Hopkins (Coursera MOOC)
  • By design a simple course, but flags some important ideas like being clear on association, inference, prediction.

Data Science in Real Life

  • Link and my Certificate
  • Johns Hopkins (Coursera MOOC)
  • Part of the same specalisation as the other JH courses, and I’d view this as part B of A Crash Course in Data Science.

Executive Data Science Capstone

  • Link and my Certificate
  • Johns Hopkins (Coursera MOOC)
  • The capstone to the JH courses. A surreal video based pick your own adventure. I learnt nothing from this, but I enjoyed the weirdness.

Other

Diversity and inclusion in the workplace

  • Link and my Certificate
  • ESSEC Business School (Coursera MOOC)
  • This is an important topic, but I feel it’s more a topic where you should understand the underlying issues - this MOOC annoyed me a bit on it’s reliance on remembering jargon rather than concepts (I guess it’s hard to assess a course on a soft-skills topic like this via multi-choice). The biggest learnings for me came from reviewing other peoples submissions for one of the modules - as hearing people explaining their own examples in their own words helped me to understand other perspectives better.
James Black
James Black
PhD (Cantab)

James Black. Kiwi | Epidemiologist | Data Scientist | Engineering enthusiast.

comments powered by Disqus