Machine Learning最新文献

Traditional statistical learning theory relies on the assumption that data are identically and independently distributed (i.i.d.). However, this assumption often does not hold in many real-life applications. In this survey, we explore learning scenarios where examples are dependent and their dependence relationship is described by a dependency graph, a commonly utilized model in probability and combinatorics. We collect various graph-dependent concentration bounds, which are then used to derive Rademacher complexity and stability generalization bounds for learning from graph-dependent data. We illustrate this paradigm through practical learning tasks and provide some research directions for future work. To our knowledge, this survey is the first of this kind on this subject.

A combination of tools and methods known as "graph mining" is used to evaluate real-world graphs, forecast the potential effects of a given graph’s structure and properties for various applications, and build models that can yield actual graphs that closely resemble the structure seen in real-world graphs of interest. However, some graph mining approaches possess scalability and dynamic graph challenges, limiting practical applications. In machine learning and data mining, among the unique methods is graph embedding, known as network representation learning where representative methods suggest encoding the complicated graph structures into embedding by utilizing specific pre-defined metrics. Co-occurrence graphs and keyword searches are the foundation of search engine optimizations for diverse real-world applications. Current work on keyword searches on graphs is based on pre-established information retrieval search criteria and does not provide semantic linkages. Recent works on co-occurrence and keyword search methods function effectively on graphs with only one layer instead of many layers. However, the graph neural network has been utilized in recent years as a branch of graph model due to its excellent performance. This paper proposes an Effective Keyword Search Co-occurrence Multi-Layer Graph mining method by employing two core approaches: Multi-layer Graph Embedding and Graph Neural Networks. We conducted extensive tests using benchmarks on real-world data sets. Considering the experimental findings, the proposed method enhanced with the regularization approach is substantially excellent, with a 10% increment in precision, recall, and f1-score.

Good models require good training data. For overparameterized deep models, the causal relationship between training data and model predictions is increasingly opaque and poorly understood. Influence analysis partially demystifies training’s underlying interactions by quantifying the amount each training instance alters the final model. Measuring the training data’s influence exactly can be provably hard in the worst case; this has led to the development and use of influence estimators, which only approximate the true influence. This paper provides the first comprehensive survey of training data influence analysis and estimation. We begin by formalizing the various, and in places orthogonal, definitions of training data influence. We then organize state-of-the-art influence analysis methods into a taxonomy; we describe each of these methods in detail and compare their underlying assumptions, asymptotic complexities, and overall strengths and weaknesses. Finally, we propose future research directions to make influence analysis more useful in practice as well as more theoretically and empirically sound.

Machine learning models always make a prediction, even when it is likely to be inaccurate. This behavior should be avoided in many decision support applications, where mistakes can have severe consequences. Albeit already studied in 1970, machine learning with rejection recently gained interest. This machine learning subfield enables machine learning models to abstain from making a prediction when likely to make a mistake. This survey aims to provide an overview on machine learning with rejection. We introduce the conditions leading to two types of rejection, ambiguity and novelty rejection, which we carefully formalize. Moreover, we review and categorize strategies to evaluate a model’s predictive and rejective quality. Additionally, we define the existing architectures for models with rejection and describe the standard techniques for learning such models. Finally, we provide examples of relevant application domains and show how machine learning with rejection relates to other machine learning research areas.

Large-scale Web-based services present opportunities for improving UI policies based on observed user interactions. We address challenges of learning such policies through offline reinforcement learning (RL). Deployed in a production system for user authentication in a major social network, it significantly improves long-term objectives. We articulate practical challenges, provide insights on training and evaluation of offline RL, and discuss generalizations toward offline RL’s deployment in industry-scale applications.

Automatic term extraction (ATE) is a natural language processing task that eases the effort of manually identifying terms from domain-specific corpora by providing a list of candidate terms. In this paper, we treat ATE as a sequence-labeling task and explore the efficacy of XLMR in evaluating cross-lingual and multilingual learning against monolingual learning in the cross-domain ATE context. Additionally, we introduce NOBI, a novel annotation mechanism enabling the labeling of single-word nested terms. Our experiments are conducted on the ACTER corpus, encompassing four domains and three languages (English, French, and Dutch), as well as the RSDO5 Slovenian corpus, encompassing four additional domains. Results indicate that cross-lingual and multilingual models outperform monolingual settings, showcasing improved F1-scores for all languages within the ACTER dataset. When incorporating an additional Slovenian corpus into the training set, the multilingual model exhibits superior performance compared to state-of-the-art approaches in specific scenarios. Moreover, the newly introduced NOBI labeling mechanism enhances the classifier’s capacity to extract short nested terms significantly, leading to substantial improvements in Recall for the ACTER dataset and consequentially boosting the overall F1-score performance.

Abstract

High dimensional learning is data-hungry in general; however, many natural data sources and real-world learning problems posses some hidden low-complexity structure that permit effective learning from relatively small sample sizes. We are interested in the general question of how to discover and exploit such hidden benign traits when problem-specific prior knowledge is insufficient. In this work, we address this question through random projection’s ability to expose structure. We study both compressive learning and high dimensional learning from this angle by introducing the notions of compressive distortion and compressive complexity. We give user-friendly PAC bounds in the agnostic setting that are formulated in terms of these quantities, and we show that our bounds can be tight when these quantities are small. We then instantiate these quantities in several examples of particular learning problems, demonstrating their ability to discover interpretable structural characteristics that make high dimensional instances of these problems solvable to good approximation in a random linear subspace. In the examples considered, these turn out to resemble some familiar benign traits such as the margin, the margin distribution, the intrinsic dimension, the spectral decay of the data covariance, or the norms of parameters—while our general notions of compressive distortion and compressive complexity serve to unify these, and may be used to discover benign structural traits for other PAC-learnable problems.

Futbol Club Barcelona operates the largest stadium in Europe (with a seating capacity of almost one hundred thousand people) and manages recurring sports events. These are influenced by multiple conditions (time and day of the week, weather, adversary) and affect city dynamics—e.g., peak demand for related services like public transport and stores. We study fine grain audience entrances at the stadium segregated by visitor type and gate to gain insights and predict the arrival behavior of future games, with a direct impact on the organizational performance and productivity of the business. We can forecast the timeline of arrivals at gate level 72 h prior to kickoff, facilitating operational and organizational decision-making by anticipating potential agglomerations and audience behavior. Furthermore, we can identify patterns for different types of visitors and understand how relevant factors affect them. These findings directly impact commercial and business interests and can alter operational logistics, venue management, and safety.

Recently, several types of neural operators have been developed, including deep operator networks, graph neural operators, and Multiwavelet-based operators. Compared with these models, the Fourier neural operator (FNO), a physics-inspired machine learning method, is computationally efficient and can learn nonlinear operators between function spaces independent of a certain finite basis. This study investigated the bounding of the Rademacher complexity of the FNO based on specific group norms. Using capacity based on these norms, we bound the generalization error of the model. In addition, we investigate the correlation between the empirical generalization error and the proposed capacity of FNO. We infer that the type of group norm determines the information about the weights and architecture of the FNO model stored in capacity. The experimental results offer insight into the impact of the number of modes used in the FNO model on the generalization error. The results confirm that our capacity is an effective index for estimating generalization errors.

Counterfactual explanations modify the feature values of an instance in order to alter its prediction from an undesired to a desired label. As such, they are highly useful for providing trustworthy interpretations of decision-making in domains where complex and opaque machine learning algorithms are utilized. To guarantee their quality and promote user trust, they need to satisfy the faithfulness desideratum, when supported by the data distribution. We hereby propose a counterfactual generation algorithm for mixed-feature spaces that prioritizes faithfulness through k-justification, a novel counterfactual property introduced in this paper. The proposed algorithm employs a graph representation of the search space and provides counterfactuals by solving an integer program. In addition, the algorithm is classifier-agnostic and is not dependent on the order in which the feature space is explored. In our empirical evaluation, we demonstrate that it guarantees k-justification while showing comparable performance to state-of-the-art methods in feasibility, sparsity, and proximity.