What role in Data Science/Analytics should you apply for?
This post is in continuation of topic Data science: The big picture and why you should care
Let’s go back to one of the confusions we had earlier
Areas: Statistics, Machine learning, AI, Business Intelligence, Decision Sciences, Data Science, Marketing Analytics, Customer Analytics, Deep Learning and what not.
The interesting (and unfortunate) thing is – even companies don’t know what they want!
When you try to do your own research, you see these complicated Venn diagrams. And you end up being more confused than you were to begin with. So, we will NOT look at these complicated diagrams.
So what are statistics, ML?
The objective of both is “learning from data“.
Statistics is a sub field of mathematics, concerned with making sense of data. In fact, we are surrounded by statistics (remember the ‘30% chances of rain’ example) and are dealing with statistics every time we are making sense of numerical information. Statistics is NOT limited to making some statistical model to predict something. Statistics has been around for ages!
Machine learning is a relatively new field (compared to statistics). ML is a sub field of computer science, where the objective is somewhat refined to making the machine/computer learn from the data. With this, the approach becomes a little different, where scalability becomes important, and coding takes a much bigger role. Algorithms are core to ML, and are these black boxes that make the prediction, and you generally put in more data to make the prediction better.
A statistician is concerned not only with the numbers obtained, but also about things like ‘significance’, ‘confidence’ and robustness of the methods. An ML practitioner on the other hand is usually concerned with accuracy in the task at hand, and it’s generalization on new data.
Statistics and ML are closely related in the sense that both try to learn from data. While in statistics, the human interpretation and judgement is central, especially when dealing with uncertainty. It’s a very broad and developed field. In the relatively new field of Machine learning, the focus is on making the computer learn, creating an algorithm that makes predictions on the data. With this, coding and scalability become paramount.
So next time you see someone throwing around these words …
… you know not to get carried away or confused, and know exactly what to think.