A reddit data science user recently asked:
“I’d like to start messing around with data science on my own. However, I know I don’t have the skill set I need yet to do that, and I’m unsure what that might actually be, anyways. Nor am I sure what the best way to get my hands dirty might actually be. Anyone have advice?”
A key to advice for this person are the phrases “messing around” and “getting my hands dirty”.
This sounds like a person who is OK with playing around, having fun, getting a feel for the lay of the data science land.
And that is exactly the attitude that is going to lead to optimal learning.
POWER OF PROJECTS
At datascience.university we favor focusing on projects, over courses, as a productive way to enhance learning about data science. And what is a “project” but something you are interested in, that you want to do, and are prepared to play around with and see what happens?
So what advice would I give this Reddit user?
Find a simple interesting project to begin to learn a few things about data science.
WHERE DO YOU START?
In my view you can get a great head start into data science by starting with Chapter 3 “Data visualisation” of “R for Data Science” by Garrett Grolemund & Hadley Wickham. This This chapter explains you how to visualize your data using ggplot2.
Data visualization is a central topic and skill in data and lends itself naturally to most, if not all, data science projects.
WHAT’S A GOOD PROJECT?
The answer’s clear: anything that interests you and is likely to both interest other people and to generalize.
For example, Jorge Fernandes’ project on visualizing police calls in Brockton, Massachusetts, is a great visualization project, and it generalizes to other towns and cities anywhere.
Data Science for Social Good has a number of neat projects, including:
- Predicting risk of long-term unemployment. The project aims to support Câmara Municipal de Cascais in understanding patterns of unemployment in the municipality and to develop a recommendation system to identify individuals with the lowest skills gap from the current job market and the type of the intervention required to bridge the gap, to increase the likelihood of people switching from unemployed to employed status. This could be extended to other communities. Why not the entire globe?
- Preventing juvenile interactions with the criminal justice system. DSSG worked with City of Milwaukee officials on applying a data resource towards intervening with at-risk youth and reducing juvenile crime. Again- extend this to your local community, and across your state or country.
MAKE A CHOICE & GET GOING!
Your choice of project has to reflect your interest: a project is not something imposed on you by someone else – that ‘s a task, or chore.
A project is something you want to do. But be smart about it: think what impact completion of your project will have on others.
LET PEOPLE KNOW
When learning data science it is most constructive and productive to share your project journey with others as you go.
Professionally a really good way to do this is on LinkedIn’s Data Mining, Statistics, Big Data, Data Visualization, and Data Science group.
Jorge Fernandes got TWO great jobs doing this, so take his lead and let others know what you are doing, and how you are doing.