Arm 401-A4: Data-Driven Decision Making Exam Questions (52 Terms) With Verified Definitions Updated 2024

Explain why a predictive model must be complex but not too complex {Ans: A predictive model must be complex enough to be accurate. However, if the model has too much complexity (in other words, it is overfit to the training data), it will not be accurate beyond the training data.}Unstructured data {Ans: Unstructured data is not organized into defined fields and is not consistent in format. Prospect notes are an example of internal unstructured data. Although the notes may be contained in a customer relationship management (CRM) database, they're not likely to be categorized or consistent from one sales rep to aterm-4nother. Unstructured external data includes information from the internet, such as social media sites. Risk managers have to use various techniques to gather, categorize, and analyze unstructured data.}Velocity {Ans: }5 Step Process for Data-Driven Decision Making {Ans: 1. Defining the risk management problem 2. Gathering quality data 3. Analyzing and modeling the problem (descriptive or predictive) 4. Determining insights: by identifying trends, relationships, behavior, and events 5. Making decisions}Purpose of a data governance program {Ans: To establish standards and oversee the management of an organization's data assets. Big data is a valuable resource, but