The Analytics' Edge-M-

In the big data landscape, it becomes imperative to understand what tools are available to gather data, to subsequently aggregate data into information and how to use this information to make better decisions. To this end, we discuss approaches to predictive analytics, which refers to the ability of discerning patterns from past and current data to prediction of future events. In particular, we focus on algorithms to tackle classification and regression problems as well as unsupervised learning techniques such as clustering and dimensionality reduction. Examples include healthcare applications such as heart attack pattern detection and financial applications such as fraud detection in regards to credit card usage. We then turn to prescriptive analytics which involves optimization of resources based on existing data and recognized patterns. Applications include revenue management at airlines and hotels or online dating sites where regression and optimization is supposed to figure out the perfect match. The analyses will be carried out in R.

Descriptive analytics:

Visualization of data patterns; predictive analytics: regression techniques, clustering methods, tree-based regression, random forests and support vector machines; prescriptive analytics: linear and integer programming