Among techniques for data analysis and automation, machine learning has received strong attention the last few years. This is mainly due to the recent breakthroughs within deep learning, but has quite rightfully renewed interest also in more simple and approachable techniques.

In this 2-day hands-on course we cover what machine learning is, gets familiar with the fundamental concepts and see examples of how it works in SAS.

OUTCOME
Learn what machine learning is
Be introduced to the theoretical concepts and their practical consequences
Get practical experience on how to work with ML in SAS software

Who should attend

Formats available	Standard duration
Classroom:	2 days

The participants are expected to have a strong quantitative background in mathematics, statistics, physics or similar. No special knowledge with respect to SAS programming is required.

This course addresses SAS Enterprise Miner software.

• What is machine learning
• Overview of the History – from Arthur Samuel to deep learning
• Difference to neighboring fields of research
• Life cycle and retraining

• Open EM and navigate around
• Getting the core data sets ready

• Overfitting
• Regularization in general

• Decision Trees
• kNN
• Neural Networks
• Support Vector Machines

• Decision threshold, Accuracy, confusion matrix
• ROC, Cumulative Lift, Cumulated Response,
• Unbalanced classes
• Interval variable measures: MSE, pm20, percentiles, ..
• The noise-to-bias illusion

• Learning curves
• Where does error come from?
• Bias Variance Trade-off

• Cross validation & leave-one-out
• Ensemble learning
• Bootstrap aggregation (aka Bagging)
• Boosting
• Gradient Boosting
• Random Forrest

• Deep Learning & Cognitive computing
• Ethics and machine learning
• Literature and suggested references