Expected Improvement (EI) is arguably the most popular acquisition function in Bayesian optimization and has found countless successful applications, but its performance is often exceeded by that of more recent methods. Notably, EI and its variants, including for the parallel and multi-objective settings, are challenging to optimize because their acquisition values vanish numerically in many regions. This difficulty generally increases as the number of observations, dimensionality of the search space, or the number of constraints grow, resulting in performance that is inconsistent across the literature and most often sub-optimal. Herein, we propose LogEI, a new family of acquisition functions whose members either have identical or approximately equal optima as their canonical counterparts, but are substantially easier to optimize numerically. We demonstrate that numerical pathologies manifest themselves in "classic" analytic EI, Expected Hypervolume Improvement (EHVI), as well as their constrained, noisy, and parallel variants, and propose corresponding reformulations that remedy these pathologies. Our empirical results show that members of the LogEI family of acquisition functions substantially improve on the optimization performance of their canonical counterparts and surprisingly, are on par with or exceed the performance of recent state-of-the-art acquisition functions, highlighting the understated role of numerical optimization in the literature.

The expected improvement (EI) is a popular technique to handle the tradeoff between exploration and exploitation under uncertainty. This technique has been widely used in Bayesian optimization but it is not applicable for the contextual bandit problem which is a generalization of the standard bandit and Bayesian optimization. In this paper, we initiate and study the EI technique for contextual bandits from both theoretical and practical perspectives. We propose two novel EI-based algorithms, one when the reward function is assumed to be linear and the other for more general reward functions. With linear reward functions, we demonstrate that our algorithm achieves a near-optimal regret. Notably, our regret improves that of LinTS \cite{agrawal13} by a factor $\sqrt{d}$ while avoiding to solve a NP-hard problem at each iteration as in LinUCB \cite{Abbasi11}. For more general reward functions which are modeled by deep neural networks, we prove that our algorithm achieves a $\tilde{\mathcal O} (\tilde{d}\sqrt{T})$ regret, where $\tilde{d}$ is the effective dimension of a neural tangent kernel (NTK) matrix, and $T$ is the number of iterations. Our experiments on various benchmark datasets show that both proposed algorithms work well and consistently outperform existing approaches, especially in high dimensions.

Foundation models are becoming valuable tools in medicine. Yet despite their promise, the best way to leverage Large Language Models (LLMs) in complex medical tasks remains an open question. We introduce a novel multi-agent framework, named **M**edical **D**ecision-making **Agents** (**MDAgents**) that helps to address this gap by automatically assigning a collaboration structure to a team of LLMs. The assigned solo or group collaboration structure is tailored to the medical task at hand, a simple emulation inspired by the way real-world medical decision-making processes are adapted to tasks of different complexities. We evaluate our framework and baseline methods using state-of-the-art LLMs across a suite of real-world medical knowledge and clinical diagnosis benchmarks, including a comparison ofLLMs' medical complexity classification against human physicians.

An important problem in sequential decision-making under uncertainty is to use limited data to compute a safe policy, which is guaranteed to outperform a given baseline strategy. In this paper, we develop and analyze a new model-based approach that computes a safe policy, given an inaccurate model of the system's dynamics and guarantees on the accuracy of this model. The new robust method uses this model to directly minimize the (negative) regret w.r.t. the baseline policy. Contrary to existing approaches, minimizing the regret allows one to improve the baseline policy in states with accurate dynamics and to seamlessly fall back to the baseline policy, otherwise. We show that our formulation is NP-hard and propose a simple approximate algorithm. Our empirical results on several domains further show that even the simple approximate algorithm can outperform standard approaches.

FastAI has a very flexible callback system that let's you greatly customize your training process. However, some of the pre-built and useful callbacks are not as easy to find without a deep dive into the documentation and to my knowledge, aren't covered in the regular courses. A common question is thus, "how do I automatically save my best model if it happens in the middle of a training run?" and the answer is to use the SaveModelCallback. The purpose of this callback, as the name implies, is to automatically save a new'best loss' model automatically during training. At the end of training, it then conveniently loads the top model so it's ready for you to continue with.

Martin Welker is the chief executive of Axonic. Even with the support of AI frameworks like TensorFlow or OpenAI, artificial intelligence still requires deep knowledge and understanding compared to a mainstream web developer. If you have built a working prototype, you are probably the smartest guy in the room. Congratulations, you are a member of a very exclusive club. With Kaggle, you can even earn decent money by solving real-world projects.

Anytime algorithms give intelligent systems the capability to trade deliberation time for quality of results. This capability is essential for successful operation in domains such as signal interpretation, real-time diagnosis and repair, and mobile robot control. What characterizes these domains is that it is not feasible (computationally) or desirable (economically) to compute the optimal answer. This article surveys the main control problems that arise when a system is composed of several anytime algorithms. These problems relate to optimal management of uncertainty and precision. After a brief introduction to anytime computation, I outline a wide range of existing solutions to the metalevel control problem and describe current work that is aimed at increasing the applicability of anytime computation. The term anytime algorithm was coined by Dean and Boddy in the mid-1980s in the context of their work on time-dependent planning (Dean and Boddy 1988; Dean 1987).

The deployment was the result of thorough evaluation and an online trial of a number of methods, including profile-based, collaborative filtering and hybrid algorithms. Results taken a few months after deployment show that the recommender system delivered its projected benefits. Traditionally these systems have been used to recommend items to users. The work we describe in this article concerns people-to-people recommendation in an online dating context. People-to-people recommendation is different from item-to-people recommendation: interactions between people are two-way, in that an approach initiated by one person to another can be accepted, rejected or ignored.

About this course: Welcome to Course 3 - Models & Frameworks to Support Sales Planning – In this course, you'll go through a conceptual approach to selling models and frameworks. As a primary learning outcome of this course, we emphasize the improvement in the analytical competencies and skills to develop sales planning and management. And the learning process goes through the application of the models and frameworks that contribute to supporting these processes. This course is aimed at professionals who seek improvement in conceptual support to the sales planning process, especially with an emphasis on applying selling models and frameworks methodology. At this point of the Strategic Sales Management specialization, you have an excellent understanding of the integration of sales planning to the strategy of the company.

While working at Tesla, I always enjoyed talking to people after they finished a factory tour. As much as they raved about the amazing automation, gigantic presses, and hundreds of robots, the reality was they only saw half of the actual manufacturing that was taking place in the building. Unknown to most visitors, the factory's "secret" second floor built many of Tesla's battery, power electronics, and drive-train systems. It was home to some of the most advanced manufacturing and automation systems in the company. Some of the robots moved at such high speeds that their arms needed to be built from carbon fiber instead of steel.