Concept Drift — Learn free

Self-check

Questions 1–10 of 10

1. How does Concept Drift differ from Data Drift?
Question 1A. In Concept Drift the data changes within the same elements; in Data Drift the elements changeB. In Data Drift the data changes within the same elements; in Concept Drift the data elements themselves changeC. They are identicalD. Concept Drift applies to OLTP and Data Drift to OLAPE. Concept Drift is expected while Data Drift is unexpected
2. By which other names is Concept Drift known?
Question 2A. Class drift, real concept drift or posterior probability shiftB. Covariate shift, virtual drift or virtual concept driftC. Scope creep or requirements driftD. Data discrepancy or reconciliation driftE. Sudden shift or gradual shift
3. Is Concept Drift an expected or unexpected behaviour in data?
Question 3A. Mostly expectedB. Always expectedC. Mostly unexpectedD. Never occurs in practiceE. Expected only in OLAP
4. In the cold drinks example, what event constitutes the Concept Drift?
Question 4A. Five new drink flavours being addedB. The factory increasing production to 1MC. The model being retrainedD. A sudden change in weather, with a week of rain and an unexpected temperature dropE. Customers switching to a competitor
5. In the cold drinks example, what fraction of the planned 1M weekly production do people consume after the weather changes?
Question 5A. 10%B. 25%C. 50%D. 75%E. 100%
6. What must a machine learning model do after Concept Drift is detected?
Question 6A. Nothing; the drift corrects itselfB. Archive the old dataC. Switch to a different query languageD. Increase the number of processorsE. Quickly change its training data and predict the new production behaviour
7. Which of the following is NOT one of the listed types of Concept Drift?
Question 7A. Sudden DriftB. Gradual DriftC. Incremental DriftD. Reoccurring DriftE. Covariate Drift
8. What is the consequence of Concept Drift for a model's output?
Question 8A. The model totally fails to produce or predict the planned outputB. Accuracy drops slightlyC. The model runs more slowlyD. The model's data elements are deletedE. Nothing changes
9. In the cold drinks example, on which external factor does buying behaviour depend?
Question 9A. Advertising spendB. Currency exchange ratesC. WeatherD. Road trafficE. Store opening hours
10. Which pair of performances is compared when Concept Drift is evaluated?
Question 10A. Development versus production codeB. OLTP versus OLAP throughputC. Canned versus ad hoc reportsD. Tested versus actual performance of machine learning modelsE. Cloud versus on-premises cost