Report Portal

Explaining The SQL Server 2014 Analysis Services Data Mining Model Lift Chart

How do I interpret the Lift Chart found on the Mining Accuracy tab of a SQL Server 2014 Analysis Services Data Mining structure?
Solution In this tip, we will examine different lift charts produced by the data mining models for three different sets of data. The first set of data, stored in tblLiftChart1, will classify perfectly with no false positives or false negatives. The second set of data, stored in tblLiftChart2, will cause the classification algorithm to predict with 50 percent accuracy. The third set of data, stored in tblLiftChart3, will cause the classification algorithm to generate 20 percent false positives and false negatives. Each table will have 20,000 rows and we will use the Neural Network algorithm to classify the data with 30 percent of the records held out for testing. Let's begin by creating three small tables with the following T-SQL Code.

Read more...

Tags: data mining

 

2007-2015 VidasSoft Systems Inc.