The definition of Machine Learning (ML) is that the computer is capable of learning without being specifically programmed for it. Within Business Intelligence, ML is used to say something about the future and ultimately take action. Within the BI Maturity Ladder, Machine Learning only appears in the upper parts.

Another application of Machine Learning is when you actually start analysing machine data. With ML, you can say something about when a certain component fails. The action that can follow is that the part is already being produced, even before the expected failure moment.

From that example you can think of many other examples. The only condition is that there is data. That is the input for Machine Learning. Just as in any BI tool, there has to be a lot of data in order to be able to say anything about the results achieved. The more data you have, the higher the accuracy becomes.

A little lower in the BI Maturity Ladder is Predictive Analytics. Here the same applies: the more data, the better. Predictive Analytics tells something about the future with varying degrees of certainty. This can be done by only looking at one result and recognising patterns in it. But it is better to collect different data that can influence a result and then analyse what the connection is.
Combining different data is where you benefit from Predictive Analytics. There are connections in the different sources that you don't see if you don't analyse the data.

The Big Data theme has risen sharply in popularity in recent years. There is no unambiguous answer as to what it means now. At one company, the interpretation of Big Data is different than at another company. But Big Data is always a dataset that is large and often in motion. Think for example of data from trucks, machines or a web service.

Terms that are related to Big Data are therefore also Internet of Things (IoT). Because of the complexity and pluriformity of Big Data, there is not just one software solution that will provide your company with Big Data.

All disciplines of the Data Scientist are covered to make a Big Data project a success. The most important thing about Big Data is that you start now. At the moment it is not very concrete for many people. So get to work with it, so that you know what you are doing when it is indispensable.

The final step in the Data Science process is visualisation. Without visualisation, the result is simply a database with numbers. By giving that data a face, you will understand the content and you can actively do something with the data. Azure's Analytics suite also includes Power BI. It communicates seamlessly with the data from Azure ML to give an example. By thinking in terms of demand, a dashboard is created with which you, as a user, can create value.

The Data Scientist knows which visualisations he can use to give the most complete answer to the question. Because the standard pie chart is not always sufficient, Power BI offers so-called Custom Visuals. For example, a heatmap can be a graphical image of a machine, where you can see the status of components. Or a floor plan of a parking garage in which you can see where a car is parked, including the occupancy rate and average parking duration. How the data is visualised therefore strongly depends on the question that was asked in the beginning.