Pattern Recognition and Machine Learning
Citations
162 citations
Cites methods from "Pattern Recognition and Machine Lea..."
...Naïve Bayes (NB) is a powerful probabilistic classifier employing a simplified version of Bayes formula to decide on a class of a new instance [61]....
[...]
162 citations
Cites background or methods from "Pattern Recognition and Machine Lea..."
...The probability of each hidden variable q(Z l ) can be efficiently inferred by forward-backward algorithm [3]....
[...]
...To optimize over qm(Z), the standard solution [3] is log (qm(Z)) = E\m[log (P (X,θ,Z))] + const, where E\m is the expectation of log (P (X,θ,Z)) taken over all variables not in qm(Z) [3]....
[...]
...We put restrictions on the family of distributions q(Z,θ), assuming that it is factorizable into a set of distributions ([3]):...
[...]
...Variational inference is to consider a restricted family of distributions q(Z,θ) and then seek the member of this family to approximate the real posterior distribution P (Z,θ|X), in the sense that the KL divergence between these two is minimized [3]....
[...]
...Hence, we propose a Monte Carlo based approximation [3]....
[...]
162 citations
162 citations
Cites background from "Pattern Recognition and Machine Lea..."
...Recent studies have proposed the so-called Kernel-PCA [74] signature for the recognition of shapes using depth sensors [75]....
[...]
...Then the Kernel-PCA descriptor is composed of the L largest eigenvalues of the kernel matrix....
[...]
...Kernel-PCA eigenvalues are invariant to rotation and translation....
[...]
...We concatenate both Shape-DNA and Kernel-PCA to create a single 3D shape descriptor that we use for classification....
[...]
...2) 3D Cloud Signatures with Kernel-PCA: In some cases, mainly due to the lack of texture, SfM gives a reconstruction with a poor mesh resolution....
[...]
161 citations