A Divide and Conquer Algorithm for Predict+Optimize with Non-convex Problems.
Ali Ugur GulerEmir DemirovicJeffrey ChanJames BaileyChristopher LeckiePeter J. StuckeyPublished in: AAAI (2022)
Keyphrases
- high accuracy
- experimental evaluation
- learning algorithm
- probabilistic model
- preprocessing
- benchmark problems
- times faster
- significant improvement
- cost function
- segmentation algorithm
- np complete
- linear programming
- convex hull
- detection algorithm
- optimal or near optimal
- neural network
- quadratic optimization problems
- worst case
- np hard
- search space
- computational complexity
- optimal solution
- simulated annealing
- dynamic programming
- multi objective
- k means
- recognition algorithm
- iterative algorithms
- convex programming
- convex constraints
- decision trees