Why should you care about data science? Let Rafal, who specialises in data science, explain how it can help you improve your business, understand your customers and products, make your employees happier, and your own job even more satisfying, in his 34-minute video. Data science combines four approaches towards working with data with one very important, overarching concept: the scientific method of reasoning, which can guide the way in which you should run experiments before making business decisions.

Combining statistics, “big data” analytics, machine learning, and the important but mundane data wrangling, or munging, underlies every data science project. To do that well, one needs a team comprised of a domain expert who understands the business need, a data expert who can find and wrangle data, and the data scientist, who also goes by many other names, such as an analytical expert or a predictive analyst. As a team, this group usually follows a three-part, iterative process of attempting to change business, identifying and readying data, and model building and validation. Indeed, much of this process is similar to the older discipline of data mining. Data mining is generally equivalent to machine learning—the differences are subtle. It applies learning algorithms to flat data, known as cases, and they are explained in the next two modules in more detail.

A survey conducted by O’Reilly Media has consistently identified several important tools that many data scientists use, notably lead by the widespread popularity of SQL, Excel, R and Python. SQL in the form of Microsoft SQL Server, Excel with Power Query, R with RStudio and Rattle, and Azure Machine Learning are the key tools presented in this online course.

To give you a feel for how practical even a short data science project may be, Rafal discusses his recent experience of working for a bank group, who were planning to close a large number of their branches. As a result of following the process of stating the business question in terms of data and models that could be built from it, only a few days later it was discovered that a very significant, even if small, group of customers would have suffered from the proposed branch closures to the point of significantly impacting bank’s operations.

If you feel you have similar questions which you might want to find answers to in your data, please make sure to continue learning with us. We will teach you how to answer them by using a very practical application of data science in your day-to-day work.

Log in or register for free to access this content.