A framework for controllable Pareto front learning with completed scalarization functions and its applications

Open Webpage

Tran Anh Tuan, Long Phi Hoang, Dung D. Le, Tran Ngoc Thang

Published: 01 Jan 2024, Last Modified: 11 Feb 2025Neural Networks 2024EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: Highlights•Controllable PFL is based on Completed Scalarization Functions optimization problem.•A theoretical for mapping between a reference vector and the Pareto optimal solution.•The proposed framework makes a less computational cost in large scale MTL.