Quiz collegato alla guida · Medio Livello
Quiz sui modelli di miscela gaussiana
Esamina i pesi della miscela, le responsabilità soft, le strutture di covarianza e le differenze tra l'adattamento della verosimiglianza GMM e le medie k.
Domanda 1 di 8
Cosa esprimono le responsabilità del GMM in una sola osservazione?
Anteprima di ogni domanda8 domande
Le risposte e le spiegazioni vengono visualizzate mentre rispondi al quiz.
- Cosa esprimono le responsabilità del GMM in una sola osservazione?
- Posterior membership probabilities across components
- A hard component label assigned before the model is fit
- The prior mixing weights shared by every observation
- Density values normalized across observations rather than components
- Quale struttura di covarianza consente a ciascun componente di avere una propria matrice di covarianza generale?
- Spherical
- Full
- Tied
- Diagonal
- Cosa esclude l’ipotesi di covarianza diagonale all’interno di ciascun componente?
- Unequal mixture weights
- Different component means
- Off-diagonal feature covariances
- Soft responsibilities
- In cosa differisce k-means da un GMM nella guida?
- GMM assigns every point to all components equally
- K-means estimates component covariances; GMM uses only nearest centroids
- Both optimize the same likelihood and always return identical assignments
- K-means minimizes squared distances; GMM estimates a probabilistic mixture density
- Perché diverse inizializzazioni GMM possono produrre adattamenti diversi?
- The likelihood is convex, so initial values cannot matter
- The mixture likelihood is nonconvex and can have local optima
- A covariance matrix prevents mixture components from overlapping
- Adding components always improves held-out likelihood
- What tradeoff comes with full covariance matrices in higher dimensions?
- They remove the need for model selection
- They force all components to share one mean
- More parameters and greater overfitting risk with limited data
- They make every responsibility equal
- What does permuting component labels do to a fitted mixture density?
- It changes soft assignments into known labels
- It makes the covariance matrices diagonal
- It guarantees better held-out fit
- It leaves the same density when parameters and weights are permuted together
- Why should a responsibility of 0.51 not be presented as certain cluster membership?
- It leaves substantial probability on competing components
- It is a guaranteed likelihood ratio of one
- It proves the model selected too many components
- It means the observation has zero density