Machine Learning Flashcards
1
Q
types of unsupervised learning
A
density estimation - clustering - dimentionality reduction blind signal separation factor analysis gaussians (fit from data)
2
Q
types of unsupervised learning
A
density estimation - clustering - dimentionality reduction blind signal separation factor analysis
3
Q
techniques for clustering
A
k-means
expectation maximation
4
Q
problems with k-mean
A
- need to know k
- local minimum (but the general problem is NP-hard)
- high dimensionality
- lack of mathematical basis
5
Q
problems with k-means
A
- need to know k
- local minimum (but the general problem is NP-hard)
- high dimensionality
- lack of mathematical basis
6
Q
techniques for clustering
A
k-means
expectation maximization
7
Q
problems with k-means
A
- need to know k
- local minimum (but the general problem is NP-hard)
- high dimensionality
- lack of mathematical basis
8
Q
maximum likehood estimator
A
…
9
Q
expectation maximization vs. k-means
A
soft vs. hard correspondance
10
Q
maximum likehood estimator
A
Finding particular parametric values that make the observed results the most probable (given the model)