Stochastic algorithms for computing means of probability measures (Q424479): Difference between revisions
From MaRDI portal
Created a new Item |
Changed an Item |
||
Property / author | |||
Property / author: Clément Dombry / rank | |||
Normal rank | |||
Property / review text | |||
Consider a probability measure \(\mu\) on a regular geodesic ball of a Riemannian manifold \(M\) with distance \(\rho\). For \(p\geq1\), a stochastic gradient descent algorithm converging almost surely to the (unique) \(p\)-mean \(e_p\) of \(\mu\) is described (\(e_p\) minimises \(x\mapsto\int_M\rho^p(x,y)\mu(dy)\)). More precisely, a time inhomogeneous Markov chain \((X_k)\) is introduced explicitly, and it is proved that \(X_k\) converges almost surely and in \(L^2\) to \(e_p\). The speed of convergence is estimated, and an invariance principle type result is proved. The advantage with respect to a deterministic gradient descent algorithm is that it is easier to implement. | |||
Property / review text: Consider a probability measure \(\mu\) on a regular geodesic ball of a Riemannian manifold \(M\) with distance \(\rho\). For \(p\geq1\), a stochastic gradient descent algorithm converging almost surely to the (unique) \(p\)-mean \(e_p\) of \(\mu\) is described (\(e_p\) minimises \(x\mapsto\int_M\rho^p(x,y)\mu(dy)\)). More precisely, a time inhomogeneous Markov chain \((X_k)\) is introduced explicitly, and it is proved that \(X_k\) converges almost surely and in \(L^2\) to \(e_p\). The speed of convergence is estimated, and an invariance principle type result is proved. The advantage with respect to a deterministic gradient descent algorithm is that it is easier to implement. / rank | |||
Normal rank | |||
Property / reviewed by | |||
Property / reviewed by: Jean Picard / rank | |||
Normal rank | |||
Property / Mathematics Subject Classification ID | |||
Property / Mathematics Subject Classification ID: 60J22 / rank | |||
Normal rank | |||
Property / Mathematics Subject Classification ID | |||
Property / Mathematics Subject Classification ID: 58C35 / rank | |||
Normal rank | |||
Property / Mathematics Subject Classification ID | |||
Property / Mathematics Subject Classification ID: 60J05 / rank | |||
Normal rank | |||
Property / Mathematics Subject Classification ID | |||
Property / Mathematics Subject Classification ID: 60F15 / rank | |||
Normal rank | |||
Property / zbMATH DE Number | |||
Property / zbMATH DE Number: 6040289 / rank | |||
Normal rank | |||
Property / zbMATH Keywords | |||
mean | |||
Property / zbMATH Keywords: mean / rank | |||
Normal rank | |||
Property / zbMATH Keywords | |||
barycenter | |||
Property / zbMATH Keywords: barycenter / rank | |||
Normal rank | |||
Property / zbMATH Keywords | |||
probability measure | |||
Property / zbMATH Keywords: probability measure / rank | |||
Normal rank | |||
Property / zbMATH Keywords | |||
Riemannian geometry | |||
Property / zbMATH Keywords: Riemannian geometry / rank | |||
Normal rank | |||
Property / zbMATH Keywords | |||
convexity | |||
Property / zbMATH Keywords: convexity / rank | |||
Normal rank | |||
Property / zbMATH Keywords | |||
geodesic ball | |||
Property / zbMATH Keywords: geodesic ball / rank | |||
Normal rank | |||
Property / zbMATH Keywords | |||
Markov chain | |||
Property / zbMATH Keywords: Markov chain / rank | |||
Normal rank | |||
Property / zbMATH Keywords | |||
invariance principle | |||
Property / zbMATH Keywords: invariance principle / rank | |||
Normal rank |
Revision as of 22:26, 29 June 2023
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | Stochastic algorithms for computing means of probability measures |
scientific article |
Statements
Stochastic algorithms for computing means of probability measures (English)
0 references
1 June 2012
0 references
Consider a probability measure \(\mu\) on a regular geodesic ball of a Riemannian manifold \(M\) with distance \(\rho\). For \(p\geq1\), a stochastic gradient descent algorithm converging almost surely to the (unique) \(p\)-mean \(e_p\) of \(\mu\) is described (\(e_p\) minimises \(x\mapsto\int_M\rho^p(x,y)\mu(dy)\)). More precisely, a time inhomogeneous Markov chain \((X_k)\) is introduced explicitly, and it is proved that \(X_k\) converges almost surely and in \(L^2\) to \(e_p\). The speed of convergence is estimated, and an invariance principle type result is proved. The advantage with respect to a deterministic gradient descent algorithm is that it is easier to implement.
0 references
mean
0 references
barycenter
0 references
probability measure
0 references
Riemannian geometry
0 references
convexity
0 references
geodesic ball
0 references
Markov chain
0 references
invariance principle
0 references