On Biased Random Walks, Corrupted Intervals, and Learning Under Adversarial Design
This work addresses foundational issues in probability and learning theory, potentially impacting ML/AI broadly, but appears incremental as it builds on existing models.
The paper tackles fundamental problems in probability theory regarding biased random walks and interval detection under noise, applying these results to learning thresholds and intervals under a new adversarial design model, but no concrete numbers are provided.
We tackle some fundamental problems in probability theory on corrupted random processes on the integer line. We analyze when a biased random walk is expected to reach its bottommost point and when intervals of integer points can be detected under a natural model of noise. We apply these results to problems in learning thresholds and intervals under a new model for learning under adversarial design.