Stochastic Adaptive Optimization with Unreliable Inputs: A Unified Framework for High-Probability Complexity Analysis Article Swipe
Scheinberg, Katya
,
Xie, Miaolan
·
YOU?
·
· 2025
· Open Access
·
· DOI: https://doi.org/10.48550/arxiv.2511.19411
YOU?
·
· 2025
· Open Access
·
· DOI: https://doi.org/10.48550/arxiv.2511.19411
We consider an unconstrained continuous optimization problem where, in each iteration, gradient estimates may be arbitrarily corrupted with a probability greater than 1/2. Additionally, function value estimates may exhibit heavy-tailed noise. This setting captures challenging scenarios where both gradient and function value estimates can be unreliable, making it applicable to many real-world problems, which can have outliers and data anomalies. We introduce an algorithmic and analytical framework that provides high-probability bounds on iteration complexity for this setting. The analysis offers a unified approach, encompassing methods such as line search and trust region.
Related Topics To Compare & Contrast
Concepts
Mathematical optimization
Computer science
Function (biology)
Outlier
Algorithm
Term (time)
Optimization problem
Value (mathematics)
Mathematics
Gradient method
Stochastic optimization
Computational complexity theory
Line (geometry)
Bellman equation
Variation (astronomy)
Line search
Optimization algorithm
Measure (data warehouse)
Iterative method
Stochastic approximation
Estimation theory
Stochastic process
Key (lock)
Metadata
- Type
- preprint
- Landing Page
- https://doi.org/10.48550/arxiv.2511.19411
- OA Status
- green
- OpenAlex ID
- https://openalex.org/W7106653192
All OpenAlex metadata
Raw OpenAlex JSON
No additional metadata available.