Scalable Unidirectional Pareto Optimality for Multi-Task Learning with Constraints

TL;DR

This paper introduces a scalable method for learning the entire Pareto front in multi-objective optimization problems, enabling flexible trade-off selection at runtime, with demonstrated improvements in multi-task learning accuracy and efficiency.

Contribution

It proposes a novel Pareto-front learning approach using Fritz-John Conditions and double gradient descent, advancing multi-task learning and constrained optimization techniques.

Findings

Achieves accurate Pareto front approximation on synthetic benchmarks.

Demonstrates improved accuracy and efficiency in multi-task learning tasks.

Shows scalability and generalization across different problem complexities.

Abstract

Multi-objective optimization (MOO) problems require balancing competing objectives, often under constraints. The Pareto optimal solution set defines all possible optimal trade-offs over such objectives. In this work, we present a novel method for Pareto-front learning: inducing the full Pareto manifold at train-time so users can pick any desired optimal trade-off point at run-time. Our key insight is to exploit Fritz-John Conditions for a novel guided double gradient descent strategy. Evaluation on synthetic benchmark problems allows us to vary MOO problem difficulty in controlled fashion and measure accuracy vs. known analytic solutions. We further test scalability and generalization in learning optimal neural model parameterizations for Multi-Task Learning (MTL) on image classification. Results show consistent improvement in accuracy and efficiency over prior MTL methods as well as techniques from operations research.