Learning to Optimize Under Constraints with Unsupervised Deep Neural Networks

01/04/2021
by   Seyedrazieh Bayati, et al.
0

In this paper, we propose a machine learning (ML) method to learn how to solve a generic constrained continuous optimization problem. To the best of our knowledge, the generic methods that learn to optimize, focus on unconstrained optimization problems and those dealing with constrained problems are not easy-to-generalize. This approach is quite useful in optimization tasks where the problem's parameters constantly change and require resolving the optimization task per parameter update. In such problems, the computational complexity of optimization algorithms such as gradient descent or interior point method preclude near-optimal designs in real-time applications. In this paper, we propose an unsupervised deep learning (DL) solution for solving constrained optimization problems in real-time by relegating the main computation load to offline training phase. This paper's main contribution is proposing a method for enforcing the equality and inequality constraints to the DL-generated solutions for generic optimization tasks.

READ FULL TEXT

page 4

page 5

research
01/04/2023

First-order penalty methods for bilevel optimization

In this paper we study a class of unconstrained and constrained bilevel ...
research
05/27/2019

Learning to Optimize with Unsupervised Learning: Training Deep Neural Networks for URLLC

Learning the optimized solution as a function of environmental parameter...
research
06/14/2016

Learning to learn by gradient descent by gradient descent

The move from hand-designed features to learned features in machine lear...
research
01/22/2022

An Unsupervised Deep Unrolling Framework for Constrained Optimization Problems in Wireless Networks

In wireless network, the optimization problems generally have complex co...
research
06/15/2023

Neural Fields with Hard Constraints of Arbitrary Differential Order

While deep learning techniques have become extremely popular for solving...
research
02/08/2022

Teaching Networks to Solve Optimization Problems

Leveraging machine learning to optimize the optimization process is an e...
research
12/13/2017

Evolving Unsupervised Deep Neural Networks for Learning Meaningful Representations

Deep Learning (DL) aims at learning the meaningful representations. A me...

Please sign up or login with your details

Forgot password? Click here to reset