SwiftVI: Time-Efficient Planning and Learning with MDPs

Kasper Overgaard Mortensen, Konstantinos Skitsas, Andreas Pavlogiannis, Davide Mottin, Panagiotis Karras

Conference on Machine Learning and Systems 2025 · Day 3 · Session 6: Edge and Cloud Systems

In the realm of artificial intelligence and machine learning, particularly within reinforcement learning and autonomous systems, the ability to plan and make optimal decisions in complex, uncertain environments is paramount. This talk introduces **SwiftVI**, a novel approach designed to significantly enhance the efficiency of **Value Iteration (VI)**, a foundational algorithm for solving **Markov Decision Processes (MDPs)**. Presented by Kasper Overgaard Mortensen from Aarhus University, SwiftVI is a collaborative effort with the University of Copenhagen, building upon master's research exploring efficient value iteration techniques. The core motivation behind SwiftVI stems from the inherent scalability challenges of traditional Value Iteration when confronted with MDPs characterized by vast state-action spaces.

AI review

SwiftVI presents a genuinely interesting algorithmic idea — using max heaps to exploit monotonic upper bounds and do selective action updates in Value Iteration — but the talk as described here is too thin to be useful. No concrete numbers, no reproducible setup, no code, and the experimental section amounts to 'we have a lot of plots and it seems quite good.' The core insight is real, but I can't tell if this is a 10% speedup or 10x, and the write-up reads like it was generated from an abstract rather than distilled from watching someone explain a system they built.