Test
Module 2 Review
Crea una cuenta gratis para terminar este test
Guarda tu progreso y consulta tu resultado final.
¿Ya tienes cuenta? Inicia sesión
Preguntas del test (12)
1In K-means Clustering, Number of clusters formed by the algorithm are
- ATo be defined by the User
- BAutomatically determined choosing optimum number of clusters
- CInitially defined by user and then automatically scaled by the algorithm
2The Centers of the Clusters in K-Means Clustering are initialized
- AAt different points Randomly, at the start of the clustering and updated at every iteration
- BUser defines the initial cluster centers and are updated at every iteration
- CAt different points Randomly, after every iteration
- DAll clusters start at the same point initially and are updated at every iteration
3K-means Clustering is an example of
- ASupervised Learning algorithm
- BUnsupervised Learning algorithm
- CReinforcement Learning Algorithm
- DNot a machine learning algorithm
4Why transformation is necessary before Clustering
- ATo bring all variables on same scale as it uses distance metrics to assign data points to clusters
- BBecause the algorithm cannot handle large values
- CTo remove outlier points to avoid potential wrong clusters
- DTo decrease the dimensions of the variables.
5Which of the following is not a disadvantage of nonhierarchical clustering methods?
- ADoes not scale up well with large/complex data
- BMakes assumptions about shape of clusters
- CSusceptible to outliers
- DNumber of clusters need to be specified in advance
6Which of the following is not an advantage of hierarchical clustering methods?
- ANot susceptible to outliers
- BAbility to capture non-spherical clusters
- COrdering of observations has no impact on cluster solutions
- DNo need for an initial guess at number of clusters.
7Which of the following best describes a hierarchical clustering method
- AInvolves a tree-like construction process where clusters at any level of the tree are a combination of clusters below that level.
- BAssigns objects into prespecified number of clusters using a distance/similarity metric. No tree-like structure exists.
- CIterates on the clusters to maximize or minimize a criterion
- DScales up well with large/complex data
8Which of the following best describes a nonhierarchical clustering method
- AAssigns objects into prespecified number of clusters using a distance/similarity metric
- BHas a tree-like structure
- CDoes not scale well for large/complex data
- DDoes not make assumptions about the shape of clusters
9Why is Base Variable Transformation done before Clustering?
- ATo give equal importance to each variable in influencing cluster results
- BTo increase the skewness to a manageable number
- CTo increase kurtosis
- DFor the variable distribution to deviate from Normal
10Which of the following is a type of Base Variable Transformation for Numeric variables?
- AAll the above
- BScale transformation
- CShape transformation
- DNumeric to Categorical
11In which of the type(s) of Clustering methods the assignment of object to cluster is not fixed through the iteration process.
- ANonhierarchical
- BHierarchical
- CAll of the above
- DNone of the above
12In which of the type(s) of Clustering methods the ordering of observations does not have any impact on clustering solution.
- AHierarchical
- BNonhierarchical
- CAll of the above
- DNone of the above
Información
- Descripción
- Test sobre Module 2 Review, creado por Nick Feldman el 01/05/2022.
- Autor
- Nick Feldman
- Idioma
- English
- Visitas
- 0
- Historial
- Creado el 1 may 2022 · actualizado el 1 may 2022
- Valoración
- Aún sin valoraciones Valora este test