Hessian estimates for convex solutions to quadratic Hessian equation (Q1719945)

The authors prove a priori Hessian estimates for convex solutions to the Hessian equation \[ \sigma_k (D^2 u)= \sigma_k (\lambda ) =\sum_{1\le i_1 < \cdots < i_k \le n} \lambda_{i_1} \cdots \lambda_{i_k}= 1 \] with \(k = 2\). Here \(\lambda_i\)s are the eigenvalues of the Hessian \(D^2u\). The main result of this article is the following theorem Theorem 1.1. Let \(u\) be a smooth solution to \(\sigma_2 (D^2 u)= 1\) on \(B_R (0) \subset \mathbb{R}^n\) with \(D^2 u \ge \left[\delta - \sqrt{2/ [n (n-1)]}\right] I\) \, for any \(\delta >0\). Then \[ |D^2u (0)| \le g(\| Du\|_{L^\infty(B_R(0))} /R, n), \] where \(g (t,n)\) is a finite and positive function for each positive \(t\) and dimension \(n\). The authors prove Theorem 1.1 by a compactness argument.