1 point Predictive models are always predict the future;

Question 1 What is the difference between predictive and descriptive models? 1 point Predictive models are always predict the future; descriptive models focus on past events. Unlike descriptive models, there is no estimation of values for predictive models. Unlike predictive models, there is no estimation of values for descriptive models. Predictive models are generally supervised models; descriptive models are generally unsupervised. Question 2 What are some common challenges with healthcare data that often require ETL work? (Select all that apply). 1 point It is often necessary to remove duplicate records. Cross-walk of data mapping issues exist. Data differences occur through time. Many tables have many rows per patient. Question 3 Which of the following represents the common transformation processes? (Select all that apply). 1 point Select Specific Records Aggregate Numeric Data Recode Data Values Applying Grouper Models to Analytical Datasets Matching Concepts to Database Fields Question 4 What are some alternative names for co-occurrence grouping? 1 point Set Mining, Market Basket Analysis Set Mining, Association Rules Mining, and Market Basket Analysis Supervised learning, Cluster analysis Market Analysis, Predictive modeling, Unsupervised learning Question 5 What did Einstein say about approaches to science and analytics? 1 point "Every analysis should match the reality of the situation being modeled. Complexity rules." "The world is complex. Embrace it and find theories to describe it perfectly." "Model the world in all of its true beauty." "Everything should be made as simple as possible, but not simpler." Question 6 What is the best definition of classification? 1 point Classification seeks to predict which of a set of classes that individuals or observations belongs to Classification is a domain of healthcare analytics related to groupers Classification is an unsupervised method to group medical terms Classification seeks to put observations or records into granular sub-categories Question 7 Which statements accurately reflect challenges associated with analyzing temporal sequences? (Select all that apply). 1 point Analyzing event sequences is relatively easy once analysts apply the SEMMA approach. Analyzing temporal sequences often takes additional effort to transform the data into the proper structure. Analyzing event sequences is like transposing data into different analytical structures. Even if it is relatively easy for a good programmer to retrieve all the raw data and timestamps, it is much harder to analyze the data to create actionable information from the raw data. Event sequencing analytics is usually not required in healthcare - groupers solve this problem. Question 8 Which statement is the best approach when starting an ETL project? 1 point Provide a complete list of all database fields to the client - ask them to select the ones that they need. Map concepts to database fields after a preliminary data extraction has been provided to the analytical team. Before looking for specific data fields available in the database, decide which concepts are required for the analytical project. Perform statistical analyses early in the process - this will help inform the ETL extraction. Question 9 What does the acronym SEMMA mean? 1 point Study, Explore, Model, Modify, Analyze Sample, Explain, Modify, Manipulate, Assess Standardize, Explore, Model, Modify, Assess Sample, Explore, Modify, Model, Assess Question 10 What are additional names for columns? What are additional names for rows? 1 point Columns: Fields, cases Rows: Variables, features Columns: Values, attributes Rows: Records, observations Columns: Variables, features, fields, or attributes Rows: Records, instances, cases, or individuals Columns: Variables Rows: Cases