A recent Graph Neural Network (GNN) approach for learning to branch has been shown to successfully reduce the running time of branch-and-bound algorithms for Mixed Integer Linear Programming (MILP). While the GNN relies on a GPU for inference, MILP solvers are purely CPU-based. This severely limits its application as many practitioners may not have access to high-end GPUs. In this work, we ask two key questions. First, in a more realistic setting where only a CPU is available, is the GNN model still competitive? Second, can we devise an alternate computationally inexpensive model that retains the predictive power of the GNN architecture? We answer the first question in the negative, and address the second question by proposing a new hybrid architecture for efficient branching on CPU machines. The proposed architecture combines the expressive power of GNNs with computationally inexpensive multi-layer perceptrons (MLP) for branching. We evaluate our methods on four classes of MILP problems, and show that they lead to up to 26% reduction in solver running time compared to state-of-the-art methods without a GPU, while extrapolating to harder problems than it was trained on.

State-of-the-art Mixed Integer Linear Programming (MILP) solvers combine systematic tree search with a plethora of hard-coded heuristics, such as branching rules. While approaches to learn branching strategies have received increasing attention and have shown very promising results, most of the literature focuses on learning fast approximations of the \emph{strong branching} rule. Instead, we propose to learn branching rules from scratch with Reinforcement Learning (RL). We revisit the work of Etheve et al. (2020) and propose a generalization of Markov Decisions Processes (MDP), which we call \emph{tree MDP}, that provides a more suitable formulation of the branching problem. We derive a policy gradient theorem for tree MDPs that exhibits a better credit assignment compared to its temporal counterpart. We demonstrate through computational experiments that this new framework is suitable to tackle the learning-to-branch problem in MILP, and improves the learning convergence.

When making decisions, people often overlook critical information or are overly swayed by irrelevant information. A common approach to mitigate these biases is to provide decision-makers, especially professionals such as medical doctors, with decision aids, such as decision trees and flowcharts. Designing effective decision aids is a difficult problem. We propose that recently developed reinforcement learning methods for discovering clever heuristics for good decision-making can be partially leveraged to assist human experts in this design process. One of the biggest remaining obstacles to leveraging the aforementioned methods for improving human decision-making is that the policies they learn are opaque to people. To solve this problem, we introduce AI-Interpret: a general method for transforming idiosyncratic policies into simple and interpretable descriptions. Our algorithm combines recent advances in imitation learning and program induction with a new clustering method for identifying a large subset of demonstrations that can be accurately described by a simple, high-performing decision rule. We evaluate our new AI-Interpret algorithm and employ it to translate information-acquisition policies discovered through metalevel reinforcement learning. The results of three large behavioral experiments showed that the provision of decision rules as flowcharts significantly improved people's planning strategies and decisions across three different classes of sequential decision problems. Furthermore, a series of ablation studies confirmed that our AI-Interpret algorithm was critical to the discovery of interpretable decision rules and that it is ready to be applied to other reinforcement learning problems. We conclude that the methods and findings presented in this article are an important step towards leveraging automatic strategy discovery to improve human decision-making.

Giving robots the ability to operate in the real world has been, and continues to be, one of the most difficult tasks in AI research. Since 1987, researchers at Carnegie Mellon University have been investigating one such task. Their research has been focused on using adaptive, vision-based systems to increase the driving performance of the Navlab line of on-road mobile robots. This research has led to the development of a neural network system that can learn to drive on many road types simply by watching a human teacher. This article describes the evolution of this system from a research project in machine learning to a robust driving system capable of executing tactical driving maneuvers such as lane changing and intersection navigation.

Loan underwriting is the process of evaluating a loan application to determine whether the loan should be funded. The process often starts with a potential borrower walking into a branch office and requesting a loan to purchase or refinance a home. A processor asks the borrower to fill out an application, setting in motion a lengthy information-gathering process in which as many as 1500 data-element pieces will eventually be collected. This loan information includes items about the borrower's employment, income, assets, liabilities, and monthly expenses. During the process, a credit report and appraisal will be ordered from a third-party vendor.