Adding a Logistic Regression Model to the Call Center Structure (Intermediate Data Mining Tutorial)
In addition to analyzing the factors that might affect call center operations, you were also asked to recommend ways for the staff to improve their service grades. In this task, you will use the same mining structure that you used to build the neural network model used for exploration of the data, and add a mining model that will be used for creating predictions.
Both neural networks and logistic regression can be used for prediction. However, in general neural networks are considered well-suited for exploration of complex interactions, whereas logistic regression is particularly well-suited to predicting binary outcomes based on known independent variables. In this tutorial, you have already identified your target outcome (better service grade) and have learned some of the factors that are likely to influence the service grade. Therefore, logistic regression is a good choice for predicting how changes in independent variables, such as staffing and call response time, might affect the service grade.
In this lesson, you will add a new model, and customize the new model for your business questions.
To add a new mining model to the call center mining structure
In Business Intelligence Development Studio, in Solution Explorer, right-click the mining structure, Call Center, and select Open Designer.
In Data Mining Designer, click the Mining Models tab.
Click Create a related mining model.
In the New Mining Model dialog box, for Model name, type Call Center - LR. For Algorithm name, select Microsoft Logistic Regression.
The new mining model is displayed in the Mining Models tab.
To customize the logistic regression model
In the column for the new mining model, Call Center - LR, leave Fact CallCenter ID as the key.
Change the value of ServiceGrade and Level Two Operators to Predict.
These columns will be used both as input and for prediction.
When you include multiple predictable attributes in a neural network model or logistic regression model, in essence you are creating two different models within the same metadata container. This is because the algorithm creates a separate subtree for each set of predictable attributes.
Change all other columns to Input.
To specify the seed and process the models
In the Mining Model tab, right-click the column for the model named Call Center - LR, and select Set Algorithm Parameters.
In the row for the HOLDOUT_SEED parameter, click the empty cell under Value, and type 1. Click OK.
The value that you choose as the seed does not matter, as long as you use the same value for all related models.
In the Mining Models menu, select Process Mining Structure and All Models. Click Yes to deploy the updated data mining project to the server.
In the Process Mining Model dialog box, click Run.
Click Close to close the Process Progress dialog box, and then click Close again in the Process Mining Model dialog box.