Quiz
Segmentation Machine Learning
Erstelle ein kostenloses Konto, um dieses Quiz zu beenden
Speichere deinen Fortschritt und sieh dir dein Endergebnis an.
oder mit deiner E-Mail registrieren
Schon ein Konto? Anmelden
Fragen in diesem Quiz (10)
1Cluster is the only technique you can use to Segment
- ATRUE
- BFALSE
- CNO ANSWER
2Which one is correct regarding the cluster algorithms we have seen?
- ANo Answer
- BK-means algorithm allows the inclusion of categorical variables without transformation
- CSilhouette final statistic is the average of every observation's Silhouette
- DYou can select the measure of distance in K-means
- ETWO-STEP Cluster (interactive in Dataiku DSS) is purely a mathematical algorithm
3We may use a gender variable coded M/F as clustering variable in a K-MEANS
- ATRUE
- BFALSE
- CNO ANSWER
4One of the outputs available in Dataiku DSS in a Segmentation analysis is a Heat Map that shows variable importance for each segment
- ATRUE
- BFALSE
- CNO ANSWER
5SILHOUETTE is the most important validation criterion on a Segmentation
- AFALSE
- BTRUE
- CNO ANSWER
6Regarding variables that take part in any stage of the segmentation project:
- ANo Answer
- BYou may use profiling variables to describe groups once they are found
- CThere is no need to differentiate beforehand clustering and profiling variables
- DClustering variables are not the only ones that enter into the algorithm
- EProfiling variables are entered in the segmentation algorithm
7General statements
- AAt least two differences between Segmentation and Classification are 1. one is unsupervised, the other one supervised and 2. In Classification we don´t find groups
- BNo Answer
- CHeat map in Dataiku DSS shows which observations belong to which clusters
- DIf Silhouette statistic is between 0,5 and 1 we can confirm the segmentation solution is a definitive one
- ESegmentation is very much an inference technique: from the sample used we use statistics to infer results to a population
8Regarding the checking list before running a Segmentation:
- ADimension reduction is something mandatory when variables are expressed in different units
- BThe idea behind variable re-scaling is to remove the effects of diferent variable scales in segmentation
- CThere should not be any correlation whatsoever between the variables entered into the segmentation algorithm
- DThe idea behind dimension reduction is to remove the effects of variable units in segmentation
- ENo Answer
9Using cosine as distance measure produces same groups as Euclidean distance since both are mathematical calculations
- ANO ANSWER
- BTRUE
- CFALSE
10If we obtain several small clusters (less than 5% of total sample size), it is normally recommendable to merge them
- ANO ANSWER
- BTRUE
- CFALSE
Informationen
- Beschreibung
- Informática Quiz am Segmentation Machine Learning, erstellt von Angel Intriago am 12/12/2021.
- Autor
- Angel Intriago
- Sprache
- English
- Aufrufe
- 3
- Schlagwörter
- Bewertung
- Noch keine Bewertungen Bewerte dieses Quiz