Analytics Portfolio

Finding “the best” can be tricky when we can value different aspects in different ways. What is best for you might not be best for me. A way to include multiple aspects when ranking a number of “contestants” is to include weights. Get data on the number of […]

The core aim when working with the Concepts is to find the key dimensions in our argumentation or view of the object/phenomenon we are relating to. We want to avoid addressing something that has a difference, without making a distinction. If apples and oranges are different, and that […]

Applied Analysis does not have to be related to a complex situation where business value, strategic alignment, or policy considerations come into play. It can be relevant even in the most basic, everyday situation where you have to make a decision. Often you probably do not even think […]

Sometimes, we start not from a problem, but from a data set. Devices, systems, websites, and other things generate data. If we start with a data set we can (normally) not influence which variables the data contains. We must therefore adapt the Questions and Concepts to the opportunities […]

Collecting data is an important part of analysis, and an important skill of an analyst. However, sometimes we can use existing data sets, either because they already exist within our organisation, or because they are easily accessible. Working with pre-existing datasets lets us practice the other aspects of […]

Find a data set you find interesting, and perform cluster analysis on it. The data should have at least two variables with good variation. Perform k-means clustering using R, Python, or an online tool. Some online tools are also great at explaining and visualizing what happens. Try https://datatab.net/statistics-calculator/cluster […]

In a research paper, Boldt et al. (2016) uses Facebook data to predict the sales of Nike. What is the business value of this? For whom? Is there a risk that the prediction will be biased? By what? Are they successful in building a predictive model? This is […]

Machine Learning is a term used for a number of different techniques where hyper-dimensional datasets can be used without prior conceptualization and definition of the relationships between variables. The “machine” will “learn” from the data. As a result, a ML-model can be very complex and thus treated as […]

The Diamond Model of Applied Analysis describes a structured approach to analysis. How do you see this? Describe your approach to analysis, use your own words or borrow the terminology from the diamond model (approx. 2 pages). Add, remove, rephrase so that it fits your approach to analysis. […]

Find a research paper in a field you are interested in. Skip the abstract and start reading the Introduction. What Questions can you think of from the Introduction? Could you guess the aim and research question before they were explicitly stated in the text? What research questions does […]