Questions tagged [convex]
A convex set includes all points lying between any two points from the set. A convex function on such a set is a function lying below any straight line connecting two points from its graph. Convex optimization is concerned with searching for the minimum of such a function.
99
questions
2
votes
0answers
14 views
Strange constraint in optimization problem
I am trying to solve the following QP :
$$ \min_{x} \quad (1/2)\|y - Ax \|^2 _2 + \lambda \|Dx\|_1$$
where $y \in \mathbb{R}^n$, $A \in \mathbb{R}^{n\times m}$, $D \in \mathbb{R}^{l\times m}$. Assume :...
1
vote
0answers
42 views
Is a decreasing and assymptotic (for y=const) convex function always logarithmically convex? [migrated]
In short, I have a function p(x) which is defined for x>=0. The function is convex, decreasing and assymptotically approaches 0 for $x\rightarrow\infty$ . I need to know if $p''p>=(p')^2$ .
As ...
0
votes
0answers
4 views
Convex variant of Bhattacharyya coefficient
For (discrete, finite) probability distributions $P,Q$, the Bhattacharyya coefficient is $B(P,Q) := \sum_x \sqrt{P_x Q_x}$. It can be shown that this is jointly concave in $P$ and $Q$. My question is, ...
1
vote
1answer
25 views
Proof: Condition for a function being strongly convex
Let $\lambda \in \mathbb{R}_{> 0}$. A function $f : \mathbb{R}^d \to \mathbb{R}$ is $\lambda$-strongly convex, if for all $\alpha \in (0, 1)$ and all $u, v \in \mathbb{R}^d$
$$
f(\alpha u + (1 - \...
4
votes
1answer
91 views
Optimization: Convex function
Problem statement
Use the definition of convexity of a function, i.e., that for any $\boldsymbol{x}$, $\boldsymbol{y} \in \mathbb{R}^{d}$ and $\lambda \in \left [0,1 \right ]$ we have
\begin{align*}
...
2
votes
0answers
32 views
Sum of SVM hinge losses always a convex function?
The data loss function for a multi class SVM may take the following expression:
\begin{equation}
L=\frac{1}{N}\sum_{i}\sum_{j\neq y_i}\left[ \max(0,w_j^Tx_i-w_{y_i}^Tx_i+1)\right]
\end{equation}
...
2
votes
1answer
20 views
Question about Decision boundary in Logistical Regression
I am a Machine Learning newbie and studying Logistical Regression.
For the data shown above, a straight line cannot separate all the positive and negative decisions. One thing I could understand is ...
1
vote
1answer
59 views
Doubt about definition of Regret in Online convex optimization setting
In online convex optimization, the regret of an algorithm $\mathcal{A}$ as defined in Introduction to Online Convex Optimization (Page 5) is:
$$
regret_T(\mathcal{A}) = \sup_{\{f_1,...,f_T\}} \sum_{t=...
0
votes
0answers
24 views
Is logistic regression always convex?
I know that the cross-entropy loss is convex, and I've seen it stated that logistic regression is convex and has a global minimum.
e.g. here https://www.cs.cmu.edu/~mgormley/courses/10701-f16/slides/...
2
votes
1answer
53 views
Iterative solution to Gamma distribution MLE problem
I'm trying to follow the derivation for the MLE parameters of the gamma distribution in [1].
The standard approach is to derive an expression for the log likelihood, differentiate with respect to ...
1
vote
0answers
26 views
Extended Lasso problem ( two L1 norms)
I am trying to find a closed form solution for the following problem :
$$
\underset{\beta}{\operatorname{minimize}} \frac{1}{2} || y - X \beta ||^{2} + \lambda || \beta ||_{1} + \mu || \beta - \gamma |...
0
votes
0answers
11 views
Kullback-Leibler divergence from density f to density g
If $f$ and $g$ are density functions that are positive over the same region, then the Kullback-Leibler divergence from density $f$ to density $g$ is defined by:
$$KL(f,g) = E_f\left[\ln\left(\frac{f(...
2
votes
0answers
70 views
ROC Convex Hull for Model Selection
The area under the convex hull of a roc curve is by construction always "better" than its area under curve. Some curves might see more of an increase in reported auc than others. Is it a viable ...
1
vote
1answer
25 views
How to understand the sufficient condition for global optimum for a constrained optimization probelm
How to understand the sufficient condition for global optimum? From my understanding, the global optimum should be 0 instead of $\geq 0$, where does this come from?
Consider the constrained ...
2
votes
1answer
43 views
Basic question about labeling of random variables apropos of the definition of convexity
I understand the definition of convexity in a function $f: \mathbb R^d \to \mathbb R$ as the inequality for all $a,b\in\mathbb R^d$ and $0<\theta<1:$
$$\theta f(a) + (1-\theta)f(b) \geq ...
0
votes
0answers
35 views
Why does quadratic optimization produce non-stable results
Edit: posing the question in a different way
I am running a mean-variance optimization, which is a quadratic optimization problem. I run the optimization 2000 times for different levels of risk (via ...
0
votes
0answers
11 views
Is penalized logistic regression convex? [duplicate]
Is logistic regression problem penalized by elastic net penalty convex optimization problem? More specifically, I want to find out whether it is suitable for dual formulation such that the duality gap ...
1
vote
0answers
21 views
How to minimize the sum of Frobenius norm and Nuclear norm
I have to minimize an objective function of the the form :
$||X_{s} - Y_{s}D_{s}||_{F}^{2} + ||D_{s}||_{F}^{2} + ||D_{s}||_{*}^{2}$
where $||.||_{F}$ denotes the Frobenius norm and $||.||_{*}$ ...
2
votes
2answers
37 views
Optimization equivalence
Can someone help me with the step by step demonstration of the following equivalence used in SVM:
$$maximize: m = \frac{1} {\|w\|} \equiv minimize: m =\frac{1} {2}\|w\|^2 $$
I would be most grateful ...
1
vote
1answer
46 views
Strongly convex function evaluated over a mean of n points
Let f(x) be a Strongly-convex function under some m > 0.
Given two points x, y it is known that:
$$f(\frac{x + y}{2}) \leq \frac{f(x) + f(y)}{2} - \frac{1}{2^3} \cdot m \cdot ||x - y||^2$$
What is ...
2
votes
1answer
25 views
Some tricky details about PCA non-convexity
I am reading about PCA and I came up with a some contradictions. PCA based on this post, PCA optimization problem is given by:
$$\begin{aligned}
\max_{w} \quad & w^T\Sigma w\\
\textrm{s.t.} \...
0
votes
1answer
7 views
Intervals from an underdetermined nonnegative linear system
I'm working on a problem in genomics that yields the following puzzle. Let $b\in \mathbb R^I$, $t$ and $p\in \mathbb R^J$, and $s \in \mathbb R^{I\times J}$. Suppose $t,b,p$ are known. Further suppose:...
1
vote
2answers
93 views
Concavity of negative binomial GLM
I need to estimate the log-likelihood of the negative binomial regression. I mean full log-likelihood, including the dispersion parameter.
The problem is: when I start LBFGS, BFGS, or gradient ...
3
votes
1answer
98 views
Why are K-means and GMM (Gaussian Mixture Models) not suitable for discovering clusters with non-convexs shapes?
I have seen that mainly here and from a lot of resources that K-means and
Hello all!
Gaussian mixtures are not suitable for detecting clusters with non-convex shapes. I know that because both ...
1
vote
1answer
37 views
Why does the formulation of the SVM problem has the bias (something we try to optimize) as a part of the constraint?
The common formulation of the SVM problem is
$$\min_{\theta, \theta_0}\frac{1}{2}||\theta||^2$$
$$\text{ subject to: } y^{(t)}(\theta \cdot x^{(t)} + \theta_0) \geq 1, \ t=1,...,n,$$
However, it ...
5
votes
2answers
221 views
L1-regularization enforces sparsity for convex functions
I have a convex function $f \colon \mathbb{R}^n \to \mathbb{R}$ that I minimize using L1-regularization:
$\DeclareMathOperator*{\argmin}{arg\,min}$
$$
x^*=\argmin_x f(x) + \lambda ||x||_1
$$
Can I ...
1
vote
1answer
45 views
D-optimal DOE suggest repeated samples
I tried to generate a D optimal design but the design output sounds very weird to me. I have a (real) process and I`d like to explore 3 factors, but the process have a lot of constraints so I provide ...
0
votes
0answers
103 views
Why backpropagation if loss function is not convex in nature?
Backpropagation contains the method of gradient decent, which works well for convex loss functions with a global minima.
But, for training, in most of the neural network tasks, backpropagation is ...
1
vote
1answer
39 views
Online Optimization - Regret in Absolute Error
In the online convex optimization literature static regret is defined as $\sum_{t=1}^{T}\left(f_t\left(x_t\right)-f_t\left(x^*\right)\right)$ where $x^*=\arg min_{x\in\mathcal{X}}\sum_{t=1}^{T}f_t\...
1
vote
0answers
72 views
Convexity of conditional expectation
Define $g(k)\equiv\mathbb{E}(X|_{X>k})$ and assume that the probability density $f$ of $X$ is twice continuously differentiable. Is there a sufficient condition in terms of $f$ that imply that $g^{...
3
votes
1answer
262 views
Prove that a function $\ln (e^{a_1} + e^{a_2} + \cdots + e^{a_n} )$ is convex?
Define the function:
$$f(a_1, a_2, · · · , a_n) = \ln (e^{a_1} + e^{a_2} + \cdots + e^{a_n} ).$$
I want to prove that $f$ is convex. Now, to show that is a function is convex, we can take second ...
1
vote
1answer
101 views
How to recover primal problem from its dual counterpart
I am asking this from context of optimization in machine learning. We often talk about a primal problem and how this primal problem can be solved by first converting it into a dual problem (Using ...
1
vote
0answers
126 views
How positive definite Hessian approximations for SGD (e.g. Gauss-Newton) handle saddles?
For example due to symmetry of parameters, functions optimized in machine learning usually have huge number of local minima and saddles - growing exponentially with dimension.
I am trying to ...
2
votes
0answers
31 views
How to solve a non-convex with equality constraint optimization problem?
I have a non-convex optimization problem with equality constraint, I can derive the KKT conditions, but it seems just one of the KKT conditions is valid. Could you please give some advice on how to ...
1
vote
2answers
558 views
Convexity of cross entropy
I am not sure if this is a better fit for this site or mathematics.stackexchange but I've seen similar questions on here before. I'd like to know if the following is true and if so, how I could go ...
2
votes
0answers
35 views
Normal equations issue
Let $A\mathbf{x}=\mathbf{b}$ be an overdetermined system, with $A$ being an $n \times m $ full-column rank rectangular matrix.
Are these minimization problems equivalent?
$$ 1) \;\underset{\mathbf{x}...
0
votes
0answers
34 views
Coordinate descent in integer programing: when does it work?
Denote $N_i=\{0,1,\dots,\bar{n}_i\}$ and define $N=N_1\times \dots \times N_I$. I want to minimize a function $f:N\rightarrow \mathbb{R}$. For the functions $f$ that interest me, it is very easy to ...
0
votes
1answer
212 views
Is error function always assumed and convex?
While updating weights of the neural network, most of the algorithms use convex optimisation because of the reason that error is a convex function.
My doubt is that whether the convex-ness of error ...
1
vote
1answer
131 views
How to understand whether Stochastic Gradient Descent has converged?
I am using SGD to solve for MSE function. My training set is around 50K, and I am monitoring the gradient at every epoch (once a pass is completed over all the training data). I played around a lot ...
2
votes
1answer
128 views
Regularizing the inverse coefficient matrix in multivariate regression
I'd like to minimize the objective
$ \operatorname{tr}[ (Y-XR^{-1})^T (Y-XR^{-1}) ] + \lambda \sum_{ij} |R_{ij}|$
wrt to $R$ (which is $P \times P$ but non-symmetric) where $Y$ and $X$ are both $N \...
3
votes
0answers
127 views
Optimization textbooks for statistics and data analytics
Any statistical analysis, machine learning or data science involves some sort of optimization at the end of the day.
I'm looking for good linear and nonlinear optimization textbooks for self ...
25
votes
6answers
5k views
For convex problems, does gradient in Stochastic Gradient Descent (SGD) always point at the global extreme value?
Given a convex cost function, using SGD for optimization, we will have a gradient (vector) at a certain point during the optimization process.
My question is, given the point on the convex, does the ...
2
votes
0answers
131 views
Vapnik-Chervonenkis Dimension
My question has to do with the VC dimension of the class of convex polygons with $m$ vertices. A solution to this problem is given in the following: Advanced Algorithms.
Call the class of all ...
3
votes
1answer
1k views
Any optimization problem can be expressed as one with a linear objective
For any standard optimization problem(think linear programming, convex, non-convex problem), would it be possible to express the optimization problem as one with a linear objective?
3
votes
1answer
70 views
Why is global convexity not possible in higher-dimensional settings for this loss function?
I was reading this paper which discusses nonconvex penalized regression. They use the following notation:
$X\in\mathbb{R}^{n\times p}$ is a data matrix with $n$ observations in $p$ variables
$\beta\...
2
votes
0answers
35 views
Why proximal gradient is a special case of sub-gradient?
While I was learning proximal gradient descent methods, I saw many similar statements as below:
$$p=prox_f(x) \Longleftrightarrow x-p \in \partial f(p) $$
for example, ...
3
votes
1answer
741 views
What is the use of the log of the sum of exponents in machine learning
I want to understand why somebody would use log sum exponent trick.
I am reading this blog.
But I don't really understand the first paragraph.
It says
Let's say we have an n-dimensional ...
1
vote
0answers
22 views
Low-variance estimate for the mean of the sotfmax transformation of a variable
Consider a set of infintiely-differentiable convex functions real-valued functions $f_i: \mathcal X \rightarrow \mathbb R$, where $i$ varies from $1$ to $m$, and suppose
we know all the moments of $...
3
votes
1answer
162 views
L2 SVM (squared hinge) theory
The linear L2 SVM can be intuitively understood as
\begin{equation}
\text{minimize } f(\boldsymbol{w}) = \frac{1}{2} \Vert\boldsymbol{w}\Vert^2_2 + C \sum_{i=1}^m \xi_i^2 \tag{1}
\end{equation}
...
0
votes
0answers
111 views
Is cross-entropy loss for MaxEnt models convex? If so, how do I prove it?
I am trying to prove if cross-entropy loss for MaxEnt models is convex.
My first attempt at approaching this problem is to compute the Hessian matrix of second-order partial derivatives and showing ...
