1591. Efficient Usage of Machine Learning-Based Approaches to Improve the Solvability of Linear Programming Problems
Invited abstract in session TB-34: Advancements of OR-analytics in statistics, machine learning and data science 3, stream Advancements of OR-analytics in statistics, machine learning and data science.
Tuesday, 10:30-12:00Room: Michael Sadler LG10
Authors (first author is the speaker)
| 1. | Ronit Neogy
|
| Decision Sciences, Indian Institute of Management Bangalore | |
| 2. | Jitamitra Desai
|
| Decision Sciences, Indian Institute of Management Bangalore |
Abstract
Recognizing that the computational time required to solve a linear programming problem can be heavily influenced by the choice of a starting basis (referred to as "warm start"), we apply machine learning techniques to apriori decide the decision variables that are most likely to be part of the optimal basis. Using historical data, several well-known classification techniques, predicted on a detailed feature set, are utilized to determine a near-optimal basis with a negligible inference cost. Furthermore, we supplement machine learning-based prediction models with analytic methods to ensure the validity of the generated starting basis. Problem parameters and their combinations that produce the most effective feature set that can be used to generate a warm-start basis are identified, and our detailed computations show that the use of advanced machine-learning techniques can significantly improve the solvability of linear programming problems.
Keywords
- Machine Learning
- Programming, Linear
- Large Scale Optimization
Status: accepted
Back to the list of papers