Order from us for quality, customized work in due time of your choice.
Overview
You are a new business analyst in the lending division of the Financial Charm Bank. You have been asked to create a simple prediction to determine the characteristics of customers who are no longer using the banking services.
You will use the customer churn data set to predict the Exited column using the Rattle tool. You will predict a binary outcome; this is called a classification algorithm.
For this activity, you will identify which of the many columns may help predict if a customer is considering changing their bank. The result of this analysis may provide opportunities for banking leadership to take action and retain these customers.
Prompt
Create a memo with the results of your analysis along with supporting screenshots (from Rattle) that illustrate the steps taken to create this simple prediction. Also, discuss the goals, data set, and fields used in the prediction, as well as the output of the model.
In your memo to the vice president, include the following criteria:
Describe the goals, data set, and fields to use for predictive modeling.
Describe how predictive modeling can be used to determine future customer churn.
Describe the data set and fields that you will use to build your model.
Perform an initial data analysis using Rattle for the given data set and submit screenshots of the analysis.
On the Data tab, identify the input and target variables.
Identify any variables you ignored.
Provide justification for your choices.
On the Explore tab, interpret Summary Statistics to describe the data.
Develop predictive modeling activities.
On the Model tab, choose a predictive modeling approach, set the parameters, run the activities, and submit screenshots of the activities.
Justify your choices.
Reflect on the predictive churns and describe the steps used.
Discuss the performance of the model.
Review the output of the model and discuss if and how this model may be useful to predict future customer churn.
Identify if the model can be applied in the same form or potential adjustments such as model tuning are needed. Explain your reasoning.
Discuss the important columns/features used in the prediction algorithm.
For explaining the prediction algorithm, first identify the important columns/features, then discuss why they are important in relation to the given data set.
These columns/features are generated in RStudio (not Rattle) in the plots pane. Include these diagrams in your memo.
Justify if this predictive modeling approach can be useful to predict customer churn.
Order from us for quality, customized work in due time of your choice.