Advanced Analytics
What is Machine Learning?
Common Examples Spam Email Filters Optical Character Recognition Coverage Risks Recommendations Fraud Detection Intrusion Detection Predictive Maintenance
Types of Machine Learning
Machine Learning Process
Machine Learning Algorithms Classification algorithms classify data into different categories that can then be used to predict one or more discrete variables, based on the other attributes in the dataset. Regression algorithms predict one or more continuous variables, such as profit or loss, based on other attributes in the dataset. Clustering algorithms determine natural groupings and patterns in datasets and are used to predict grouping classifications for a given variable.
Azure ML
Azure ML Studio
Demo: Creating an Azure Machine Learning experiment Creating a workspace Loading historical data Visualizing the data Splitting data into training and validation sets Training the model
Evaluating Models Receiver Operator Characteristic (ROC) curves Displays the fraction of true positives out of the total actual positives. The higher and further to the left, the more accurate the model is. Precision/Recall curves Precision represents the fraction of retrieved instances that are relevant. Recall represents the fraction of relevant instances that are retrieved. Lift curves This format is a variation on the ROC curve Measures the fraction of true positives, in relation to the target response probability.
Comparing Models
Demo: Refining the Model Score the model Evaluate the model Comparing models
Demo: Publish and Consume the Model Creating the web service Testing the web service Calling RRS service from a client app Batch execution
A word from my sponsor… Data Analytics and BI specialists Six-time Microsoft Federal Partner of the Year Four-time Microsoft State and Local Government Partner of the Year Four-time Microsoft Windows Partner of the Year Data Analytics and BI specialists Microsoft Office 365 and Microsoft Cloud support SharePoint and Dynamics CRM expertise (301) 721-0100 | www.go-planet.com | info@go-planet.com