In this work, we study the weighted empirical risk minimization (weighted ERM) schema, in which an additional data-dependent weight function is incorporated when the empirical risk function is being minimized. We show that under a general ``balanceable" Bernstein condition, one can design a weighted ERM estimator to achieve superior performance in certain sub-regions over the one obtained from standard ERM, and the superiority manifests itself through a data-dependent constant term in the error bound. These sub-regions correspond to large-margin ones in classification settings and low-variance ones in heteroscedastic regression settings, respectively. Our findings are supported by evidence from synthetic data experiments.

The neural process (NP) is a family of computationally efficient models for learning distributions over functions. However, it suffers from under-fitting and shows suboptimal performance in practice. Researchers have primarily focused on incorporating diverse structural inductive biases, \textit{e.g.} attention or convolution, in modeling. The topic of inference suboptimality and an analysis of the NP from the optimization objective perspective has hardly been studied in earlier work. To fix this issue, we propose a surrogate objective of the target log-likelihood of the meta dataset within the expectation maximization framework. The resulting model, referred to as the Self-normalized Importance weighted Neural Process (SI-NP), can learn a more accurate functional prior and has an improvement guarantee concerning the target log-likelihood. Experimental results show the competitive performance of SI-NP over other NPs objectives and illustrate that structural inductive biases, such as attention modules, can also augment our method to achieve SOTA performance. Our code is available at \url{https://github.com/hhq123gogogo/SI_NPs}.

Most learning problems can be solved by minimization of log loss. This bare fact is inescapable in modern AI and machine learning - the variety is in the details. What is the space of measured data? What is the support of the distribution? Changing such properties of the problem fundamentally changes learning behavior, leading to the variety of modeling approaches successfully used in data science. But for many inference and decision-making tasks, log loss can be axiomatically inescapable. We explore such loss minimization problems in the language of statistical mechanics, which studies how systems of "particles" like atoms can be approximately described by relatively few bulk properties. There is a direct analogue to modeling, where large datasets are described by relatively few model parameters.

Traditionally, TD and FQI are viewed as differing in the number of updates toward the target value function: TD makes one update, FQI makes an infinite number, and Partial Fitted Q-Iteration (PFQI) performs a finite number, such as the use of a target network in Deep Q-Networks (DQN) in the OPE setting. This perspective, however, fails to capture the convergence connections between these algorithms and may lead to incorrect conclusions, for example, that the convergence of TD implies the convergence of FQI. In this paper, we focus on linear value function approximation and offer a new perspective, unifying TD, FQI, and PFQI as the same iterative method for solving the Least Squares Temporal Difference (LSTD) system, but using different preconditioners and matrix splitting schemes. TD uses a constant preconditioner, FQI employs a data-feature adaptive preconditioner, and PFQI transitions between the two. Then, we reveal that in the context of linear function approximation, increasing the number of updates under the same target value function essentially represents a transition from using a constant preconditioner to data-feature adaptive preconditioner. This unifying perspective also simplifies the analyses of the convergence conditions for these algorithms and clarifies many issues. Consequently, we fully characterize the convergence of each algorithm without assuming specific properties of the chosen features (e.g., linear independence). We also examine how common assumptions about feature representations affect convergence, and discover new conditions on features that are important for convergence. These convergence conditions allow us to establish the convergence connections between these algorithms and to address important questions.

The PAC-Bayesian framework has significantly advanced the understanding of statistical learning, particularly for majority voting methods. Despite its successes, its application to multi-view learning -- a setting with multiple complementary data representations -- remains underexplored. In this work, we extend PAC-Bayesian theory to multi-view learning, introducing novel generalization bounds based on R\'enyi divergence. These bounds provide an alternative to traditional Kullback-Leibler divergence-based counterparts, leveraging the flexibility of R\'enyi divergence. Furthermore, we propose first- and second-order oracle PAC-Bayesian bounds and extend the C-bound to multi-view settings. To bridge theory and practice, we design efficient self-bounding optimization algorithms that align with our theoretical results.

Existing methods to summarize posterior inference for mixture models focus on identifying a point estimate of the implied random partition for clustering, with density estimation as a secondary goal (Wade and Ghahramani, 2018; Dahl et al., 2022). We propose a novel approach for summarizing posterior inference in nonparametric Bayesian mixture models, prioritizing density estimation of the mixing measure (or mixture) as an inference target. One of the key features is the model-agnostic nature of the approach, which remains valid under arbitrarily complex dependence structures in the underlying sampling model. Using a decision-theoretic framework, our method identifies a point estimate by minimizing posterior expected loss. A loss function is defined as a discrepancy between mixing measures. Estimating the mixing measure implies inference on the mixture density and the random partition. Exploiting the discrete nature of the mixing measure, we use a version of sliced Wasserstein distance. We introduce two specific variants for Gaussian mixtures. The first, mixed sliced Wasserstein, applies generalized geodesic projections on the product of the Euclidean space and the manifold of symmetric positive definite matrices. The second, sliced mixture Wasserstein, leverages the linearity of Gaussian mixture measures for efficient projection.

Understanding the world and explaining it with scientific theories is a central aspiration of artificial intelligence research. Proposing theories, designing experiments to test them, and then revising them based on data are fundamental to scientific discovery. Despite the significant promise of LLM-based scientific agents, no benchmarks systematically test LLM's ability to propose scientific models, collect experimental data, and revise them in light of new data. We introduce BoxingGym, a benchmark with 10 environments for systematically evaluating both experimental design (e.g. collecting data to test a scientific theory) and model discovery (e.g. proposing and revising scientific theories). To enable tractable and quantitative evaluation, we implement each environment as a generative probabilistic model with which a scientific agent can run interactive experiments. These probabilistic models are drawn from various real-world scientific domains ranging from psychology to ecology. To quantitatively evaluate a scientific agent's ability to collect informative experimental data, we compute the expected information gain (EIG), an information-theoretic quantity which measures how much an experiment reduces uncertainty about the parameters of a generative model. A good scientific theory is a concise and predictive explanation. Therefore, to quantitatively evaluate model discovery, we ask a scientific agent to explain their model and then assess whether this explanation enables another scientific agent to make reliable predictions about this environment. In addition to this explanation-based evaluation, we compute standard model evaluation metrics such as prediction errors. We find that current LLMs, such as GPT-4o, struggle with both experimental design and model discovery. We find that augmenting the LLM-based agent with an explicit statistical model does not reliably improve these results.

Marketing mix modeling (MMM) is a widely used method to assess the effectiveness of marketing campaigns and optimize marketing strategies. Bayesian MMM is an advanced approach that allows for the incorporation of prior information, uncertainty quantification, and probabilistic predictions (1). In this paper, we describe the process of building a Bayesian MMM model for the online insurance company Lemonade. We first collected data on Lemonade's marketing activities, such as online advertising, social media, and brand marketing, as well as performance data. We then used a Bayesian framework to estimate the contribution of each marketing channel on total performance, while accounting for various factors such as seasonality, market trends, and macroeconomic indicators. To validate the model, we compared its predictions with the actual performance data from A/B-testing and sliding window holdout data (2). The results showed that the predicted contribution of each marketing channel is aligned with A/B test performance and is actionable. Furthermore, we conducted several scenario analyses using convex optimization to test the sensitivity of the model to different assumptions and to evaluate the impact of changes in the marketing mix on sales. The insights gained from the model allowed Lemonade to adjust their marketing strategy and allocate their budget more effectively. Our case study demonstrates the benefits of using Bayesian MMM for marketing attribution and optimization in a data-driven company like Lemonade. The approach is flexible, interpretable, and can provide valuable insights for decision-making.

In this paper we address the problem of performing Bayesian inference for the parameters of a nonlinear multi-output model and the covariance matrix of the different output signals. We propose an adaptive importance sampling (AIS) scheme for multivariate Bayesian inversion problems, which is based in two main ideas: the variables of interest are split in two blocks and the inference takes advantage of known analytical optimization formulas. We estimate both the unknown parameters of the multivariate non-linear model and the covariance matrix of the noise. In the first part of the proposed inference scheme, a novel AIS technique called adaptive target adaptive importance sampling (ATAIS) is designed, which alternates iteratively between an IS technique over the parameters of the non-linear model and a frequentist approach for the covariance matrix of the noise. In the second part of the proposed inference scheme, a prior density over the covariance matrix is considered and the cloud of samples obtained by ATAIS are recycled and re-weighted to obtain a complete Bayesian study over the model parameters and covariance matrix. ATAIS is the main contribution of the work. Additionally, the inverted layered importance sampling (ILIS) is presented as a possible compelling algorithm (but based on a conceptually simpler idea). Different numerical examples show the benefits of the proposed approaches

Many critical decisions, such as personalized medical diagnoses and product pricing, are made based on insights gained from designing, observing, and analyzing a series of experiments. This highlights the crucial role of experimental design, which goes beyond merely collecting information on system parameters as in traditional Bayesian experimental design (BED), but also plays a key part in facilitating downstream decision-making. Most recent BED methods use an amortized policy network to rapidly design experiments. However, the information gathered through these methods is suboptimal for down-the-line decision-making, as the experiments are not inherently designed with downstream objectives in mind. In this paper, we present an amortized decision-aware BED framework that prioritizes maximizing downstream decision utility. We introduce a novel architecture, the Transformer Neural Decision Process (TNDP), capable of instantly proposing the next experimental design, whilst inferring the downstream decision, thus effectively amortizing both tasks within a unified workflow. We demonstrate the performance of our method across several tasks, showing that it can deliver informative designs and facilitate accurate decision-making.