In ICE #9 “join4” we combine the join3 table and the customer master table using the variable CustID and in inner join. This results in a table with 1,168 observations. If we instead link this tables using a left join and the TerritoryID variable, we get a table with 18,937 observations. In this joined table, there are observations where the customerID from the join3 table does not equal the customerID from the customer master table. 

What critical input do Agarwal et al. argue is becoming “cheap” (i.e., reduced cost) and is relevant to this course on data analytics and machine learning?

In ICE #9, If we “full join” KMPG’s Sales Order file with Shipments file using Sales Order ID, we keep the records (observations) that are only in both Sales Order file and Shipments file.

In the KPMG revenue case (ICE #9), we first filtered observations for shipments made in 2017 and the “aggregated” sales transactions by TerritoryID, Shipping Month, and whether the sale is related to the new product (#7123).  We then outputted this data as a csv file. The grouped data has 5*12*2 = 120 observations.

Machine learning generates predictions by evaluating thousands (if not millions) of if-then statements. 

All else equal, a higher true positive rate is associated with what? Select all the apply. 

What of the following are true statements regarding select statements? Select all that apply.

Which of the following statements accurately describes a relational database? Select all that apply. 

Which of the following are true regarding joins? Select all that apply.

Data types define what kind of information is stored in a particular column (variable). What data type would you use to store a manager’s name?