In this paper, we address the problem of learning the structure of a pairwise graphical model from samples in a high-dimensional setting. Our first main result studies the sparsistency, or consistency in sparsity pattern recovery, properties of a forward-backward greedy algorithm as applied to general statistical models. As a special case, we then apply this algorithm to learn the structure of a discrete graphical model via neighborhood estimation. As a corollary of our general result, we derive sufficient conditions on the number of samples n, the maximum node-degree d and the problem size p, as well as other conditions on the model parameters, so that the algorithm recovers all the edges with high probability. Our result guarantees graph selection for samples scaling as n = Omega(d^2 log(p)), in contrast to existing convex-optimization based algorithms that require a sample complexity of \Omega(d^3 log(p)). Further, the greedy algorithm only requires a restricted strong convexity condition which is typically milder than irrepresentability assumptions. We corroborate these results using numerical simulations at the end.

In this paper, we give a new sharp generalization bound of lp-MKL which is a generalized framework of multiple kernel learning (MKL) and imposes lp-mixed-norm regularization instead of l1-mixed-norm regularization. We utilize localization techniques to obtain the sharp learning rate. The bound is characterized by the decay rate of the eigenvalues of the associated kernels. A larger decay rate gives a faster convergence rate. Furthermore, we give the minimax learning rate on the ball characterized by lp-mixed-norm in the product space. Then we show that our derived learning rate of lp-MKL achieves the minimax optimal rate on the lp-mixed-norm ball.

Cancer kills millions of people each year. From an AI perspective, finding effective treatments for cancer is a high-dimensional search problem characterized by many molecularly distinct cancer subtypes, many potential targets and drug combinations, and a dearth of high quality data to connect molecular subtypes and treatments to responses. The broadening availability of molecular diagnostics and electronic medical records, presents both opportunities and challenges to apply AI techniques to personalize and improve cancer treatment. We discuss these in the context of Cancer Commons, a “rapid learning” community where patients, physicians, and researchers collect and analyze the molecular and clinical data from every cancer patient, and use these results to individualize therapies. Research opportunities include: adaptively-planning and executing individual treatment experiments across the whole patient population, inferring the causal mechanisms of tumors, predicting drug response in individuals, and generalizing these findings to new cases. The goal is to treat each patient in accord with the best available knowledge, and to continually update that knowledge to benefit subsequent patients. Achieving this goal is a worthy grand challenge for AI.

Automatic transfer of learned knowledge from one task or domain to another offers great potential to simplify and expedite the construction and deployment of intelligent systems. In practice however, there are many barriers to achieving this goal. In this article, we present a prototype system for the real-world context of transferring knowledge of American football from video observation to control in a game simulator. We trace an example play from the raw video through execution and adaptation in the simulator, highlighting the system's component algorithms along with issues of complexity, generality, and scale. We then conclude with a discussion of the implications of this work for other applications, along with several possible improvements.

In this article, we discuss the problem of transferring search heuristics from one planner to another. More specifically, we demonstrate how to transfer the domain-dependent heuristics acquired by one planner into a second planner. Our motivation is to improve the efficiency and the efficacy of the second planner by allowing it to use the transferred heuristics to capture domain regularities that it would not otherwise recognize. Our experimental results show that the transferred knowledge does improve the second planner's performance on novel tasks over a set of seven benchmark planning domains.

In this paper we present a lossless technique to compress pattern databases (PDBs) in the Pancake Sorting problems. This compression technique together with the choice of zero-cost operators in the construction of additive PDBs reduces the memory requirement for PDBs in these problems to a great extent, thus making otherwise intractable problems able to be efficiently handled. Also, using this method, we can construct some problem-size independent PDBs. This precludes the necessity of constructing new PDBs for new problems with different numbers of pancakes. In addition to our compression technique, by maximizing over the heuristic value of additive PDBs and the modified version of the gap heuristic, we have obtained powerful heuristics for the burnt pancake problem.

Monte-Carlo Tree Search (MCTS) is remarkably successful in two-player games, but parallelizing MCTS has been notoriously difficult to scale well, especially in distributed environments. For a distributed parallel search, transposition-table driven scheduling (TDS) is known to be efficient in several domains. We present a massively parallel MCTS algorithm, that applies the TDS parallelism to the Upper Confidence bound Applied to Trees (UCT) algorithm, which is the most representative MCTS algorithm. To drastically decrease communication overhead, we introduce a reformulation of UCT called Depth-First UCT. The parallel performance of the algorithm is evaluated on clusters using up to 1,200 cores in artificial game-trees. We show that this approach scales well, achieving 740-fold speedups in the best case.

Real-time agent-centric algorithms have been used for learning and solving problems since the introduction of the LRTA* algorithm in 1990. In this time period, numerous variants have been produced, however, they have generally followed the same approach in varying parameters to learn a heuristic which estimates the remaining cost to arrive at a goal state. This short paper discusses the history and implications of learning g-costs, both alone and in conjunction with learning h-costs as an introduction to the new f-LRTA* algorithm which learns both.

This paper presents an approach to differentially constrained robot motion planning and efficient re-planning. Satisfaction of differential constraints is guaranteed by the search space which consists of motions that satisfy the constraints by construction. Any systematic replanning algorithm, e.g. D*, can be utilized to search the state lattice to find a motion plan that satisfies the differential constraints, and to repair it efficiently in the event of a change in the environment. Further efficiency is obtained by varying the fidelity of representation of the planning problem. High fidelity is utilized where it matters most, while it is lowered in the areas that do not affect the quality of the plan significantly. The paper presents a method of modifying the fidelity between replans, thereby enabling dynamic flexibility of the search space, while maintaining its compatibility with replanning algorithms. The approach is especially suited for mobile robotics applications in unknown challenging environments. We successfully applied the motion planner on a real robot: the planner featured 10Hz average replan rate on minimal computing hardware, while satisfying the car-like differential constraints.

In a number of graph search-based planning problems, the value of the cost function that is being minimized also affects the set of possible actions at some or all the states in the graph. In such planning problems, the cost function typically becomes one of the state variables thereby increasing the dimensionality of the planning problem, and consequently the size of the graph that represents the problem. In this paper, we show how to avoid this increase in the dimensionality for weighted search (with bounded suboptimality) whenever the availability of the actions is monotonically non-increasing with the increase in the cost function.