scientific article; zbMATH DE number 4031421
[For the entire collection see Zbl 0627.00013.] Let \((M,g)\) be a complete finite dimensional Riemannian manifold. Suppose we have a real function \(f: M\to R\) and we want to find one of its minima. The remark that -grad f is a vector field orthogonal to the hypersurface of constant level of f, which shows at every point \(x\in M\) the direction of steepest descent suggests that the most suitable solution of the inequality \(df(x_ i)(X)<0\) is \(X_ i=-\text{grad} f(x_ i)\). From this choice results an iterative process which we call ``the method of steepest descent of the gradient method, having in mind the well-known Euclidean model.
- Riemannian gradient algorithm for the numerical solution of linear matrix equations
- Cayley-transform-based gradient and conjugate gradient algorithms on Grassmann manifolds
- Riemannian gradient algorithm for the numerical solution of Stein equations
- On the convergence of gradient descent for finding the Riemannian center of mass
- scientific article; zbMATH DE number 938816 (Why is no real title available?)
- scientific article; zbMATH DE number 6177815 (Why is no real title available?)
- Convergence Analysis of Gradient Algorithms on Riemannian Manifolds without Curvature Constraints and Application to Riemannian Mass
- Stochastic Gradient Descent on Riemannian Manifolds
- Steepest descent method on a Riemannian manifold: the convex case
- On Gradient Flows Initialized near Maxima
- Discretized gradient flow for manifold learning
This page was built for publication:
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3771986)