Week four has passed at the Data School (time flies, doesn’t it?) and a lot is going on. Next week we have our first meeting with a client (Virgin Train Atlantic) and the Tableau exam is in two weeks.

This week we have learnt a lot, as always. We mainly focused on Alteryx and we discovered a bit more about the potential of the software. I especially took interest in Machine Learning. There are two main machine learning algorithms: Supervised Learning and Unsupervised Learning. With the Supervised Learning we have an idea of the kind of relationship between the input subject and the desired output, this way we already know how our output should look like. In other words, we are given two set of data, the input with which we need to find the best predicted model for your desired output.

Machine Supervised learning can be classified into Prediction and Classification. They are both used to predict something and while Prediction, estimate a value from a continuous set, Classification output variable takes discrete values. Listed here are the main models you can use in Alteryx to conduct the study:

  • Prediction:

Linear Regression

Spline

Boosted Model

Gamma

  • Classification:

Boosted Model

Logistic Regression (only for binary outcome)

Decision Tree

Forest Model

Naïve

Once you run the models (after choosing the most appropriate variables) you can pick the best one and do your prediction! 😊 A special thanks to Bene who came to Data School to teach us that day!