arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2510.25366cs.LGcs.AImath.OC
收藏

一种面向深度神经网络的凸性依赖两阶段训练算法

A Convexity-dependent Two-Phase Training Algorithm for Deep Neural Networks

Institute of Computer Science, University of St.Gallen (HSG)(施普伦格尔大学计算机科学学院);Institute of Computer Science in Vorarlberg, University of St.Gallen (HSG)(沃斯堡大学计算机科学研究所)

Tomas Hrycej, Bernhard Bermeitinger, Massimo Pavone, Götz-Henrik Wiegand, Siegfried Handschuh

AI总结:

该研究提出基于损失函数从非凸向凸转变假设的两阶段训练算法,用Adam和CG分别对应不同阶段,实验证实可显著提升深度神经网络的收敛性与准确率。

摘要:

机器学习的核心任务是最小化衡量模型对训练数据拟合程度的损失函数,高效实现该目标的数值方法依赖于损失函数的性质,其中最具决定性的性质是损失函数的凸性或非凸性。损失函数可能存在且经常存在非凸区域,这一事实使得Adam等非凸方法被广泛采用。然而,局部最小值意味着在其周围的某些环境中函数是凸的,在该环境中,共轭梯度(Conjugate Gradient,CG)等二阶最小化方法可保证超线性收敛。我们基于真实世界任务中的损失函数会从初始非凸状态向最优值附近的凸状态转变的假设,提出了一个新颖框架。我们利用这一特性设计了一种创新的两阶段优化算法,该算法通过观察梯度范数对损失的依赖关系来检测转变点,在对应区域分别使用非凸算法(Adam)和凸算法(CG)。计算实验证实了以下假设:这种简单的凸性结构足够常见,可在实际中被利用以显著提升收敛性和准确率。

英文摘要:

The key task of machine learning is to minimize the loss function that measures the model fit to the training data. The numerical methods to do this efficiently depend on the properties of the loss function. The most decisive among these properties is the convexity or non-convexity of the loss function. The fact that the loss function can have, and frequently has, non-convex regions has led to a widespread commitment to non-convex methods such as Adam. However, a local minimum implies that, in some environment around it, the function is convex. In this environment, second-order minimizing methods such as the Conjugate Gradient (CG) give a guaranteed superlinear convergence. We propose a novel framework grounded in the hypothesis that loss functions in real-world tasks swap from initial non-convexity to convexity towards the optimum. This is a property we leverage to design an innovative two-phase optimization algorithm. The presented algorithm detects the swap point by observing the gradient norm dependence on the loss. In these regions, non-convex (Adam) and convex (CG) algorithms are used, respectively. Computing experiments confirm the hypothesis that this simple convexity structure is frequent enough to be practically exploited to substantially improve convergence and accuracy.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑