As hyper-parameters are ubiquitous and can significantly affect the model performance, hyper-parameter optimization is extremely important in machine learning.

Cubic-regularized Newton's method (CR) is a popular algorithm that guarantees

As hyper-parameters are ubiquitous and can significantly affect the model performance, hyper-parameter optimization is extremely important in machine learning. In this paper, we consider a sub-class of hyper-parameter optimization problems, where the hyper-gradients are not available. Such problems frequently appear when the performance metric is non-differentiable or the hyper-parameter is not continuous. However, existing algorithms, like Bayesian optimization and reinforcement learning, often get trapped in local optimals with poor performance. To address the above limitations, we propose to use cubic regularization to accelerate convergence and avoid saddle points. First, we adopt stochastic relaxation, which allows obtaining gradient and Hessian information without hyper-gradients. Then, we exploit the rich curvature information by cubic regularization. Theoretically, we prove that the proposed method can converge to approximate second-order stationary points, and the convergence is also guaranteed when the lower-level problem is inexactly solved. Experiments on synthetic and real-world data demonstrate the effectiveness of our proposed method.

Cubic-regularized Newton's method (CR) is a popular algorithm that guarantees to produce a second-order stationary solution for solving nonconvex optimization problems. However, existing understandings of convergence rate of CR are conditioned on special types of geometrical properties of the objective function. In this paper, we explore the asymptotic convergence rate of CR by exploiting the ubiquitous Kurdyka-Lojasiewicz (KL) property of the nonconvex objective functions. In specific, we characterize the asymptotic convergence rate of various types of optimality measures for CR including function value gap, variable distance gap, gradient norm and least eigenvalue of the Hessian matrix. Our results fully characterize the diverse convergence behaviors of these optimality measures in the full parameter regime of the KL property. Moreover, we show that the obtained asymptotic convergence rates of CR are order-wise faster than those of first-order gradient descent algorithms under the KL property.

Cubic-regularized Newton's method (CR) is a popular algorithm that guarantees

As hyper-parameters are ubiquitous and can significantly affect the model performance, hyper-parameter optimization is extremely important in machine learning. In this paper, we consider a sub-class of hyper-parameter optimization problems, where the hyper-gradients are not available. Such problems frequently appear when the performance metric is non-differentiable or the hyper-parameter is not continuous. However, existing algorithms, like Bayesian optimization and reinforcement learning, often get trapped in local optimals with poor performance. To address the above limitations, we propose to use cubic regularization to accelerate convergence and avoid saddle points.

As hyper-parameters are ubiquitous and can significantly affect the model performance, hyper-parameter optimization is extremely important in machine learning. In this paper, we consider a sub-class of hyper-parameter optimization problems, where the hyper-gradients are not available. Such problems frequently appear when the performance metric is non-differentiable or the hyper-parameter is not continuous. However, existing algorithms, like Bayesian optimization and reinforcement learning, often get trapped in local optimals with poor performance. To address the above limitations, we propose to use cubic regularization to accelerate convergence and avoid saddle points.

The paper investigates the convergence of different measurement of cubic regularization method for non-convex optimization under KL property. It consists with a list of work on CR methods based on the analysis of Nesterove el.s' work. Since the type of methods can guarantee the convergence to the second-order stationary point, it is quite popular also considering the raising of training neural networks. The paper is well-written, clear-organized and the theorems and proofs are easy to follow. Note that this is a pure theoretical work i.e., without new algorithms and/or numerical experiments.

In this paper, we study second-order algorithms for the convex-concave minimax problem, which has attracted much attention in many fields such as machine learning in recent years. We propose a Lipschitz-free cubic regularization (LF-CR) algorithm for solving the convex-concave minimax optimization problem without knowing the Lipschitz constant. It can be shown that the iteration complexity of the LF-CR algorithm to obtain an $\epsilon$-optimal solution with respect to the restricted primal-dual gap is upper bounded by $\mathcal{O}(\frac{\rho\|z^0-z^*\|^3}{\epsilon})^{\frac{2}{3}}$, where $z^0=(x^0,y^0)$ is a pair of initial points, $z^*=(x^*,y^*)$ is a pair of optimal solutions, and $\rho$ is the Lipschitz constant. We further propose a fully parameter-free cubic regularization (FF-CR) algorithm that does not require any parameters of the problem, including the Lipschitz constant and the upper bound of the distance from the initial point to the optimal solution. We also prove that the iteration complexity of the FF-CR algorithm to obtain an $\epsilon$-optimal solution with respect to the gradient norm is upper bounded by $\mathcal{O}(\frac{\rho\|z^0-z^*\|^2}{\epsilon})^{\frac{2}{3}}$. Numerical experiments show the efficiency of both algorithms. To the best of our knowledge, the proposed FF-CR algorithm is the first completely parameter-free second-order algorithm for solving convex-concave minimax optimization problems, and its iteration complexity is consistent with the optimal iteration complexity lower bound of existing second-order algorithms with parameters for solving convex-concave minimax problems.