This course teaches you how to optimize the performance of predictive models beyond the basics by implementing various data munging and wrangling techniques. The course continues the development of supervised learning models that begins in the Machine Learning Using SAS Viya course and extends it to ensemble modeling. Running unsupervised learning and semi-supervised learning models is also discussed. In this course, you learn how to do feature engineering and clustering of variables, and how to preprocess nominal variables and detect anomalies. This course uses Model Studio, the pipeline flow interface in SAS Viya that enables you to prepare, develop, compare, and deploy advanced analytics models. Importing and running external models in Model Studio is also discussed, including open-source models. SAS Viya automation capabilities at each level of machine learning are also demonstrated, followed by some tips and tricks with Model Studio.

The self-study e-learning includes:

• Annotatable course notes in PDF format.
• Virtual lab time to practice.

Learn how to

• Develop a series of supervised learning models based on techniques such as logistic regression, decision tree, neural network, and support vector machine.
• Evaluate classifier performance of your model.
• Create an ensemble model based on different techniques.
• Preprocess and engineer features from categorical and continuous data to improve the performance of your machine learning models.
• Extract features using principal component analysis, singular value decomposition, robust principal component analysis, autoencoders, and variable clustering.
• Discover the basic concepts of cluster analysis, and then study a set of typical clustering methodologies, algorithms, and applications.
• Use statistics and machine learning to detect anomalies in your data.
• Implement a semi-supervised learning model.
• Import and run SAS 9 models in Model Studio.
• Run open-source models in Model Studio.
• Automate different stages of machine learning in SAS Viya.
• Generate automated pipelines using REST API.

Who should attend

Advanced machine learning modelers who use Model Studio

Formats available	Standard duration
Classroom:	3 days
Live Web Classroom:	6 half-day session(s)	System Requirements
e-Learning:	21 hours/180 day license	System Requirements

Before attending this course, it is recommended that you have done the following:

• Completed the Machine Learning Using SAS Viya course.
• Obtained some experience with creating and managing SAS data sets, which you can gain from the SAS Programming 1: Essentials course.
• Acquired some experience building statistical models using SAS Visual Data Mining and Machine Learning software.

This course addresses SAS Visual Statistics, SAS Visual Data Mining and Machine Learning software.

Machine Learning Fundamentals

• Model Studio review.
• Classifier performance.
• Ensemble learning.

Feature Engineering

• Introduction to feature engineering.
• Principal component analysis.
• Singular value decomposition.
• Robust principal component analysis.
• Autoencoders.
• Transforming categorical variables.

Clustering of Variables and Observations

• Variable clustering.
• Cluster analysis.

Anomaly Detection

• Introduction to anomaly detection.
• Support vector data description.
• Semi-supervised learning.

External Models in Model Studio

• Importing SAS Enterprise Miner models.
• Running SAS/STAT or SAS Enterprise Miner models.
• Running open-source models.

Machine Learning Automation

• Automation in SAS Viya.
• Data preprocessing and feature engineering.
• Modeling.
• Automated pipeline creation.
• Pipeline automation using REST API (self-study).

Tips and Tricks with Model Studio

• Managing metadata.
• Working with analysis elements.
• Using the SAS Code node.
• Interpreting models with extracted features.
• Scoring unsupervised learning models.