Association and Apriori Execution Test - Machine Learning

1. 1ASSOCIATION analysis is a valuable tool to interpret causal relationship of events

   1. ARichtig
   2. BFalsch
2. 2Antecedents & Consequents should be marked at the beginning of any ASSOCIATION analysis

   1. ARichtig
   2. BFalsch
3. 3Main pillars of association analysis are Frequency of itemsets and Reliability of rules found

   1. ARichtig
   2. BFalsch
4. 4Antecedent support is its relative frequency, also called Prior Confidence

   1. ARichtig
   2. BFalsch
5. 5“Lift” measures the ratio of rule’s confidence to the overall support of antecedents

   1. ARichtig
   2. BFalsch
6. 6Using confidence ratio as rule selection criterion can help to find interesting rules with low prior confidence

   1. ARichtig
   2. BFalsch
7. 7Using confidence ratio as rule selection can help us to find lots of trivial rules for very common consequents

   1. ARichtig
   2. BFalsch
8. 8Rule support is always lower or equal than Support

   1. ARichtig
   2. BFalsch
9. 9For a market basket analysis, tabular data normally requires specifying a customer field and dummy product variables

   1. ARichtig
   2. BFalsch
10. 10APRIORI algorithm allows multiple consequents

    1. ARichtig
    2. BFalsch

11. 11Transactional data is the preferred format for customer-level predictions

    1. ARichtig
    2. BFalsch
12. 12Approximately 5.6% of tickets contain yogurt and whole milk

    

    1. ARichtig
    2. BFalsch
13. 1340.31% of tickets containing whole milk include also tropical fruit

    

    1. ARichtig
    2. BFalsch
14. 141,372 tickets verify rule #20

    

    1. ARichtig
    2. BFalsch
15. 15Buyers of root_vegetables have bought approx. 2.25 more other vegetables than the average buyers of other vegetables

    

    1. ARichtig
    2. BFalsch
16. 16Less than 25% of customers bought “other vegetables”

    

    1. ARichtig
    2. BFalsch
17. 1710.9% of customers bought “other vegetables”

    

    1. ARichtig
    2. BFalsch
18. 18If we set Confidence Difference lower bound to 60, we would not get rules from root_vegetables

    

    1. ARichtig
    2. BFalsch

Beschreibung

Informática Quiz am Association and Apriori Execution Test - Machine Learning, erstellt von Angel Intriago am 12/12/2021.

Autor

Angel Intriago

Sprache

English

Aufrufe

10

Schlagwörter

informática

Bewertung

Noch keine Bewertungen

Bewerte dieses Quiz

Bewerte dieses Quiz

Mehr zu Association Rules and Frequent Patterns