Data and Decision Science Network Meeting – Machine Learning Taster

Join this general presentation for anyone curious about what machine learning is (and isn’t). Held on 7 October 2021, at 12.00 PM AEDT, this talk aims to give an “in a nutshell” introduction to some of the most commonly used/mentioned techniques in machine learning (ML), while also making the connection to methods in general and statistical techniques. The importance of the trade-offs of variability and bias, while presenting complex concepts in a basic manner to explain what the methods do, and that ML involves more than just “clicking a button”, will be showcased using simple one-on-one variable examples.

Today, new technologies are generating measurements for almost everything with the ability to store these massive amounts of variables and data. There is a need to exploit such data; hence, analyses are evolving and are aiming at finding the complex relations in the data. There are many techniques to approach analysis, but some techniques are in much more “fashion” and people are wanting to use them. This is the case of “machine learning” techniques, which are sometimes referred to as if the term “machine learning” was ONE specific technique, and this is observed repeatedly in grants and proposals, where phrases like “we will use machine learning to…” are used, when in fact such a term is an umbrella term for many potential techniques. Given this, it is of importance that researchers get at least a taste of what “machine (statistical) learning” is and what it is not, and what is gained/lost in the use of some of these techniques, plus knowing that it is not a magical “one click” solution.

Learn more and register here.