{"id":"arxiv/2608.05680","url":"https://arxiv.org/abs/2608.05680","tags":["math.OC"],"score":0,"title":"Curvature Residual Geometry in Bregman Regression","domain":"math.OC","source":"arxiv","authors":["Vu Khac Ky"],"reviews":0,"subject":"math.OC","summary":"We consider linear regression models fitted by minimizing Bregman losses of the form \\[ \\frac{1}{n}\\sum_{i=1}^n \\left[ φ(y_i)-φ(x_i^\\topθ) -φ'(x_i^\\topθ) \\bigl(y_i-x_i^\\topθ\\bigr) \\right], \\] where \\(y_i\\) is the observed response and \\(x_i^\\topθ\\) is the linear prediction. Even when the generating potential \\(φ\\) is strongly convex, the resulting regression objective may be nonconvex in \\(θ\\). This creates a gap between the convexity of the generating potential and the optimization geometry of the fitted model. The Hessian can be written as a weighted Gram matrix whose weights depend on the…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-08-06","reactions":{}}