Data analytics is an increasingly popular field in constant movement. We recognise the words, and we have a notion of its meaning, but there is a lot of room for interpretation. On top of that, recent evolutions in the field only add to the existing confusion around this ‘buzz word’. By taking a closer look at the recent developments we can clear up some clouds of confusion, and give you some tips to ensure you know exactly what people mean when they talk about data analytics. After all, the most satisfying part of my job is helping clients find meaning!
It is important to understand what people mean when they say they’re experts in something. Machine learning is a great example of this, as this term covers a wide array of meanings. It’s an umbrella term covering self-learning algorithms – which is what the term refers to in its most narrow sense – as well as all kinds of predictive modelling techniques. As a consequence, claiming to be an expert in machine learning does not really give us much information. It is always useful to look deeper – our job is not only to give you answers, but also to help you ask the right questions.
Broadly speaking, there are two types of data science: one area focuses on obtaining the most accurate algorithms and predictions, using the latest software and methodologies. The other focuses on obtaining the most valuable insights from the data, insights that can be transformed into tangible plans and actions. At this point, there are not many data scientists who are experts in both and allow you to enjoy the benefits of each type. It is long-term experience that makes the difference in identifying the story behind the data. I’m proud that every day, my colleagues use their experience with passion and meet each challenge head on!
Paid software used to be staples in our field, but we are now moving towards open source software like Python and R, to name just two from the expanding market. The greatest advantage open source software offers is the strong community behind it. We can all learn from each other and share code and experience. This allows for swift development and a smaller learning curve for those starting out with data science. For me, it’s important that data scientists use what they are comfortable with, as the results are largely the same regardless of which software you use.
As data scientists, it is our mission to provide answers to your business questions. During this search for answers, it can be tempting to reach for the most complex methods. We believe that the simpler the model is, the more easily it can be implemented and understood. Using the most complex methods may sound exciting, doing so when simpler methods can give you the same results will only lead to confusion. Only when less complex methods cannot provide answers, we move on to the next level.
It is our task and our pleasure to keep a close eye on all developments in data analytics, and we will keep you updated via this blog. If you want to get in touch you can drop us a line on our contact page!