# Can we use reinforcement learning and convex optimization to solve an optimization problem?

Operations Research Asked by qinqinxiaoguai on August 19, 2021

For an optimization problem, there are multiple-type variables should be optimized. Can we use the convex optimization method to solve a subproblem of partial variables, and then, with the obtained results of the subproblem, solve the remaining subproblem of other variables by reinforcement learning?

You can use RL in any step. But problem is optimality check of solution which is explained above. Also you can solve your problem directly using RL such as RL for VRP. And you can read this blog which is about RL usage. By the way your question is too broad to give detailed answer.

Answered by kur ag on August 19, 2021

Not really, but approximately. By OR standards, a problem is "solved" once we manage to satisfy the KKT conditions. There is no machine learning algorithm to date that can consistently satisfy constraints. ML is designed to give pretty good approximations, and that's about it.

For instance, image recognition can be posed as a convex optimisation problem, and we all know how well ML works on those problems. That doesn't mean however that it will always work (and it doesn't), unlike an optimisation algorithm.

Answered by Nikos Kazazakis on August 19, 2021

## Related Questions

### Conditions required for strong duality to hold for SDPs

1  Asked on August 19, 2021

### Relationship between extreme points and optimal solutions of SDPs

1  Asked on August 19, 2021

### How to parallelize metaheuristics algorithms (Island Model)?

1  Asked on August 19, 2021 by antarctica

### Can we get closed form solution for such a problem?

1  Asked on August 19, 2021

### Can we use reinforcement learning and convex optimization to solve an optimization problem?

2  Asked on August 19, 2021 by qinqinxiaoguai

### Nonlinear integer (0/1) programming solver

6  Asked on March 1, 2021 by rajya

### How can I find the shortest path for all nodes in a graph from a source $s$?

1  Asked on March 1, 2021 by windbreeze

### Free solver for MINP problems

1  Asked on February 18, 2021 by dspinfinity

### Where I can study some job shop scheduling by course (video )?

0  Asked on February 18, 2021 by yue-chao

### Linear objective function with power term in constraint

1  Asked on February 15, 2021 by user152503

### Formulating these logical constraint in an ILP

1  Asked on January 18, 2021

### Modeling the multiplication of two binary decision variables in undirected graph in python

0  Asked on January 18, 2021 by amedeo

### Flexible Job Shop with Preemption

0  Asked on January 15, 2021 by robert-hildebrand

### How to handle an equality constraint in metaheuristic algorithms (like GA, PSO)?

3  Asked on January 11, 2021 by stevgates

### What is the difference between min- cut formulation and (bi) partitioning formulation?

1  Asked on January 8, 2021 by fathese

### Logical constraint in ILP

1  Asked on December 22, 2020 by che

### Quasi-convex function must be “partially monotonic”?

1  Asked on December 13, 2020 by high-gpa

### Constraint programming resources

3  Asked on November 28, 2020 by joffrey-l

### Pyomo variable creation dilemma

1  Asked on October 31, 2020 by ethan-deakins

### Convexity of the variance of a mixture distribution

1  Asked on September 25, 2020 by independentvariable