Source author record

Baocheng Zhu

Explore this author’s publications, collaborators and research topics. This catalog record has not been claimed by its author.

ResearcherUnclaimed source record

Machine Learning math.MG math.DG math.FA Artificial Intelligence math.AP

Catalog footprint

Research overview

6works

6topics

4close collaborators

Actions

Is this your work?

Open graph Browse papers

preprint2022arXiv

On the $L_p$ Brunn-Minkowski theory and the $L_p$ Minkowski problem for $C$ -coconvex sets

Let $C$ be a pointed closed convex cone in $\mathbb{R}^n$ with vertex at the origin $o$ and having nonempty interior. The set $A\subset C$ is $C$ -coconvex if the volume of $A$ is finite and $A^{\bullet}=C\setminus A$ is a closed convex set. For $0<p<1$ , the $p$ -co-sum of $C$ -coconvex sets is introduced, and the corresponding $L_p$ Brunn-Minkowski inequality for $C$ -coconvex sets is established. We also define the $L_p$ surface area measures, for $0\neq p\in \mathbb{R}$ , of certain $C$ -coconvex sets, which are critical in deriving a variational formula of the volume of the Wulff shape associated with a family of functions obtained from the $p$ -co-sum. This motivates the $L_p$ Minkowski problem aiming to characterize the $L_p$ surface area measures of $C$ -coconvex sets. The existence of solutions to the $L_p$ Minkowski problem for all $0\neq p\in \mathbb{R}$ is established. The $L_p$ Minkowski inequality for $0<p<1$ is proved and is used to obtain the uniqueness of the solutions to the $L_p$ Minkowski problem for $0<p<1$ . For $p=0$ , we introduce $(1-τ)\diamond A_1\oplus_0τ\diamond A_2$ , the log-co-sum of two $C$ -coconvex sets $A_{1}$ and $A_{2}$ with respect to $τ\in(0, 1)$ , and prove the log-Brunn-Minkowski inequality of $C$ -coconvex sets. The log-Minkowski inequality is also obtained and is applied to prove the uniqueness of the solutions to the log-Minkowski problem that characterizes the cone-volume measures of $C$ -coconvex sets. Our result solves an open problem raised by Schneider in [Schneider, Adv. Math., 332 (2018), pp. 199-219].

preprint2020arXiv

A Riemannian Primal-dual Algorithm Based on Proximal Operator and its Application in Metric Learning

In this paper, we consider optimizing a smooth, convex, lower semicontinuous function in Riemannian space with constraints. To solve the problem, we first convert it to a dual problem and then propose a general primal-dual algorithm to optimize the primal and dual variables iteratively. In each optimization iteration, we employ a proximal operator to search optimal solution in the primal space. We prove convergence of the proposed algorithm and show its non-asymptotic convergence rate. By utilizing the proposed primal-dual optimization technique, we propose a novel metric learning algorithm which learns an optimal feature transformation matrix in the Riemannian space of positive definite matrices. Preliminary experimental results on an optimal fund selection problem in fund of funds (FOF) management for quantitative investment showed its efficacy.

preprint2020arXiv

Neural Physicist: Learning Physical Dynamics from Image Sequences

We present a novel architecture named Neural Physicist (NeurPhy) to learn physical dynamics directly from image sequences using deep neural networks. For any physical system, given the global system parameters, the time evolution of states is governed by the underlying physical laws. How to learn meaningful system representations in an end-to-end way and estimate accurate state transition dynamics facilitating long-term prediction have been long-standing challenges. In this paper, by leveraging recent progresses in representation learning and state space models (SSMs), we propose NeurPhy, which uses variational auto-encoder (VAE) to extract underlying Markovian dynamic state at each time step, neural process (NP) to extract the global system parameters, and a non-linear non-recurrent stochastic state space model to learn the physical dynamic transition. We apply NeurPhy to two physical experimental environments, i.e., damped pendulum and planetary orbits motion, and achieve promising results. Our model can not only extract the physically meaningful state representations, but also learn the state transition dynamics enabling long-term predictions for unseen image sequences. Furthermore, from the manifold dimension of the latent state space, we can easily identify the degree of freedom (DoF) of the underlying physical systems.

preprint2020arXiv

Riemannian Proximal Policy Optimization

In this paper, We propose a general Riemannian proximal optimization algorithm with guaranteed convergence to solve Markov decision process (MDP) problems. To model policy functions in MDP, we employ Gaussian mixture model (GMM) and formulate it as a nonconvex optimization problem in the Riemannian space of positive semidefinite matrices. For two given policy functions, we also provide its lower bound on policy improvement by using bounds derived from the Wasserstein distance of GMMs. Preliminary experiments show the efficacy of our proposed Riemannian proximal policy optimization algorithm.

preprint2016arXiv

The Orlicz-Petty bodies

This paper is dedicated to the Orlicz-Petty bodies. We first propose the homogeneous Orlicz affine and geominimal surface areas, and establish their basic properties such as homogeneity, affine invariance and affine isoperimetric inequalities. We also prove that the homogeneous geominimal surface areas are continuous, under certain conditions, on the set of convex bodies in terms of the Hausdorff distance. Our proofs rely on the existence of the Orlicz-Petty bodies and the uniform boundedness of the Orlicz-Petty bodies of a convergent sequence of convex bodies. Similar results for the nonhomogeneous Orlicz geominimal surface areas are proved as well.

preprint2013arXiv

The mixed $L_p$ geominimal surface areas for multiple convex bodies

In this paper, we introduce several mixed $L_p$ geominimal surface areas for multiple convex bodies for all $p\neq -n$ . Our definitions are motivated from an equivalent formula for the mixed $p$ -affine surface area. Some properties, such as the affine invariance, for these mixed $L_p$ geominimal surface areas are proved. Related inequalities, such as, Alexander-Fenchel type inequality, Santaló style inequality, affine isoperimetric inequalities, and cyclic inequalities are established. Moreover, we also study some properties and inequalities for the $i$ -th mixed $L_p$ geominimal surface areas for two convex bodies.

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

Baocheng Zhu

Research overview

Is this your work?

6 publications

On the $L_p$ Brunn-Minkowski theory and the $L_p$ Minkowski problem for $C$ -coconvex sets

A Riemannian Primal-dual Algorithm Based on Proximal Operator and its Application in Metric Learning

Neural Physicist: Learning Physical Dynamics from Image Sequences

Riemannian Proximal Policy Optimization

The Orlicz-Petty bodies

The mixed $L_p$ geominimal surface areas for multiple convex bodies

See the researcher in context

Building this map preview

Baocheng Zhu

Research overview

Is this your work?

6 publications

On the LpL_pLp​ Brunn-Minkowski theory and the LpL_pLp​ Minkowski problem for CCC-coconvex sets

A Riemannian Primal-dual Algorithm Based on Proximal Operator and its Application in Metric Learning

Neural Physicist: Learning Physical Dynamics from Image Sequences

Riemannian Proximal Policy Optimization

The Orlicz-Petty bodies

The mixed LpL_pLp​ geominimal surface areas for multiple convex bodies

See the researcher in context

Building this map preview

On the $L_p$ Brunn-Minkowski theory and the $L_p$ Minkowski problem for $C$ -coconvex sets

The mixed $L_p$ geominimal surface areas for multiple convex bodies