Random Function Descent
–Neural Information Processing Systems
Classical worst-case optimization theory neither explains the success of optimization in machine learning, nor does it help with step size selection. In this paper we demonstrate the viability and advantages of replacing the classical'convex function' framework with a'random function' framework.
Neural Information Processing Systems
May-25-2025, 17:53:53 GMT
- Country:
- Europe (0.92)
- North America
- Canada > Alberta (0.14)
- United States
- Arizona (0.14)
- Massachusetts (0.14)
- New Jersey (0.14)
- New York (0.14)
- Genre:
- Research Report > Experimental Study (1.00)
- Industry:
- Education > Educational Setting > Online (0.92)
- Technology: