Consider a function ##f\in\mathcal{C}^2## with Lipschitz continuous gradient (with constant ##L##)- we also assume the function is lowerbounded and has at least one minimum. Let ##\{x^k\}_k## be the sequence generated by Gradient Descent algorithm with initial point $x^0$ and step-size...
Consider a function $f\in\mathcal{C}^2$ with Lipschitz continuous gradient (with constant $L$)- we also assume the function is lowerbounded and has at least one minimum. Let $\{x^k\}_k$ be the sequence generated by Gradient Descent algorithm with initial point $x^0$ and step-size $0<\alpha<2/L$...
Two KL functions $f_1:\mathbb{R}^n\rightarrow \mathbb{R}$ and $f_2:\mathbb{R}^n\rightarrow \mathbb{R}$ are given which have KL exponent $\alpha_1$ and $\alpha_2$. What is the KL exponent of $f_1+f_2$?
Yes, that is true. But I am looking for more general class of function and a well-established theory on the conditions and properties of such functions. This is a bit similar to holder conditions but on the inverse of ##f##.
Let ##f:\mathbb{R}^n\rightarrow\mathbb{R}^n##. Is there any class of function and some type of "growth conditions" such that bounds like below can be established:
\begin{equation}
||f(x)||\geq g\left( \text{dist}(x,\mathcal{X})\right),
\end{equation}
with ##\mathcal{X}:= \{x:f(x)=0\}## (zero...
Hello! I have the following question:
Let $f:\mathbb{R}^n\rightarrow \mathbb{R}^n$. Is there any class of class of functions and some kind of "growth conditions" such that bounds like below can be established:
\begin{equation}
||f(x)||\geq g\left( \text{dist}(x,\mathcal{X})\right)...
Let f:\mathbb{R}^m\rightarrow\mathbb{R}^m. Define the zero set by \mathcal{Z}\triangleq\{x\in\mathbb{R}^m | f(x)=\mathbf{0}\} and an \epsilon-approximation of this set by \mathcal{Z}_\epsilon\triangleq\{x\in\mathbb{R}^m|~||f(x)||\leq\epsilon\} for some \epsilon>0. Clearly \mathcal{Z}\subseteq...