MIPaaL: Mixed Integer Program as a Layer

Machine learning components commonly appear in larger decision-making pipelines; however, the model training process typically focuses only on a loss that measures accuracy between predicted values and ground truth values. Decision-focused learning explicitly integrates the downstream decision problem when training the predictive model, in order to optimize the quality of decisions induced by the predictions. It has been successfully applied to several limited combinatorial problem classes, such as those that can be expressed as linear programs (LP), and submodular optimization. However, these previous applications have uniformly focused on problems from specific classes with simple constraints. Here, we enable decision-focused learning for the broad class of problems that can be encoded as a Mixed Integer Linear Program (MIP), hence supporting arbitrary linear constraints over discrete and continuous variables. We show how to differentiate through a MIP by employing a cutting planes solution approach, which is an exact algorithm that iteratively adds constraints to a continuous relaxation of the problem until an integral solution is found. We evaluate our new end-to-end approach on several real world domains and show that it outperforms the standard two phase approaches that treat prediction and prescription separately, as well as a baseline approach of simply applying decision-focused learning to the LP relaxation of the MIP.

[1]  Sanjeeb Dash,et al.  Lattice-free sets, multi-branch split disjunctions, and mixed-integer programming , 2014, Math. Program..

[2]  Tatsuya Akutsu,et al.  IPknot: fast and accurate prediction of RNA secondary structures with pseudoknots using integer programming , 2011, Bioinform..

[3]  Yi-Hao Kao,et al.  Directed Regression , 2009, NIPS.

[4]  Lise Getoor,et al.  Collective Classification in Network Data , 2008, AI Mag..

[5]  Lantao Yu,et al.  Deep Reinforcement Learning for Green Security Game with Online Information , 2018, AAAI Workshops.

[6]  Gautam Kaul,et al.  An Anatomy of Trading Strategies , 1998 .

[7]  J. Zico Kolter,et al.  OptNet: Differentiable Optimization as a Layer in Neural Networks , 2017, ICML.

[8]  Milind Tambe,et al.  Melding the Data-Decisions Pipeline: Decision-Focused Learning for Combinatorial Optimization , 2018, AAAI.

[9]  Le Song,et al.  2 Common Formulation for Greedy Algorithms on Graphs , 2018 .

[10]  Max Welling,et al.  Attention, Learn to Solve Routing Problems! , 2018, ICLR.

[11]  Sanjeeb Dash,et al.  Cutting planes from extended LP formulations , 2017, Math. Program..

[12]  He He,et al.  Learning to Search in Branch and Bound Algorithms , 2014, NIPS.

[13]  Andrew L. Maas Rectifier Nonlinearities Improve Neural Network Acoustic Models , 2013 .

[14]  David Manlove,et al.  Position-Indexed Formulations for Kidney Exchange , 2016, EC.

[15]  Yair Zick,et al.  Diversity Constraints in Public Housing Allocation , 2017, AAMAS.

[16]  Vipin Kumar,et al.  A Fast and High Quality Multilevel Scheme for Partitioning Irregular Graphs , 1998, SIAM J. Sci. Comput..

[17]  Priya L. Donti,et al.  SATNet: Bridging deep learning and logical reasoning using a differentiable satisfiability solver , 2019, ICML.

[18]  Lawrence V. Snyder,et al.  Reinforcement Learning for Solving the Vehicle Routing Problem , 2018, NeurIPS.

[19]  George L. Nemhauser Integer Programming: the Global Impact , 2013 .

[20]  Andrea Lodi,et al.  On learning and branching: a survey , 2017 .

[21]  Le Song,et al.  Learning to Branch in Mixed Integer Programming , 2016, AAAI.

[22]  Hamed Mohsenian Rad,et al.  Optimal Residential Load Control With Price Prediction in Real-Time Electricity Pricing Environments , 2010, IEEE Transactions on Smart Grid.

[23]  Yoshua Bengio,et al.  Machine Learning for Combinatorial Optimization: a Methodological Tour d'Horizon , 2018, Eur. J. Oper. Res..

[24]  Yoshua Bengio,et al.  Using a Financial Training Criterion Rather than a Prediction Criterion , 1997, Int. J. Neural Syst..

[25]  Ashish Sabharwal,et al.  Guiding Combinatorial Optimization with UCT , 2012, CPAIOR.

[26]  Louis Wehenkel,et al.  A Supervised Machine Learning Approach to Variable Branching in Branch-And-Bound , 2014 .

[27]  Alain Guinet,et al.  NURSE SCHEDULING USING INTEGER LINEAR PROGRAMMING AND CONSTRAINT PROGRAMMING , 2006 .

[28]  Dimitris Bertsimas,et al.  Portfolio Construction Through Mixed-Integer Programming at Grantham, Mayo, Van Otterloo and Company , 1999, Interfaces.

[29]  Lantao Yu,et al.  Deep Reinforcement Learning for Green Security Games with Real-Time Information , 2018, AAAI.

[30]  Adam N. Elmachtoub,et al.  Smart "Predict, then Optimize" , 2017, Manag. Sci..

[31]  Sergey Ioffe,et al.  Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift , 2015, ICML.

[32]  Milind Tambe,et al.  Decision-Focused Learning of Adversary Behavior in Security Games , 2019, ArXiv.

[33]  Egon Balas,et al.  Gomory cuts revisited , 1996, Oper. Res. Lett..

[34]  R. Gomory AN ALGORITHM FOR THE MIXED INTEGER PROBLEM , 1960 .

[35]  David B. Shmoys,et al.  Data Analysis and Optimization for (Citi)Bike Sharing , 2015, AAAI.

[36]  Luca Antiga,et al.  Automatic differentiation in PyTorch , 2017 .

[37]  J. Zico Kolter,et al.  What game are we playing? End-to-end learning in normal and extensive form games , 2018, IJCAI.

[38]  James Bailey,et al.  Predict+Optimise with Ranking Objectives: Exhaustively Learning Linear Functions , 2019, IJCAI.

[39]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[40]  Nitish Srivastava,et al.  Dropout: a simple way to prevent neural networks from overfitting , 2014, J. Mach. Learn. Res..

[41]  Milind Tambe,et al.  End-to-End Game-Focused Learning of Adversary Behavior in Security Games , 2020, AAAI.

[42]  Nicholas Bambos,et al.  Optimizing Intensive Care Unit Discharge Decisions with Patient Readmissions , 2012, Oper. Res..

[43]  Joshua B. Tenenbaum,et al.  End-to-End Differentiable Physics for Learning and Control , 2018, NeurIPS.

[44]  Maria-Florina Balcan,et al.  Learning to branch: Generalization guarantees and limits of data-independent discretization , 2018, ICML.

[45]  D. Michael Warner,et al.  Scheduling Nursing Personnel According to Nursing Preference: A Mathematical Programming Approach , 1976, Oper. Res..

[46]  George L. Nemhauser,et al.  Learning to Run Heuristics in Tree Search , 2017, IJCAI.

[47]  Laurence A. Wolsey,et al.  Integer and Combinatorial Optimization , 1988 .

[48]  William D'haeseleer,et al.  Effect of the accuracy of price forecasting on profit in a Price Based Unit Commitment , 2010 .

[49]  W. Karush Minima of Functions of Several Variables with Inequalities as Side Conditions , 2014 .