We propose new graph kernels grounded in the study of metric graphs via tropical algebraic geometry. In contrast to conventional graph kernels that are based on graph combinatorics such as nodes, edges, and subgraphs, our graph kernels are purely based on the geometry and topology of the underlying metric space. A key characterizing property of our construction is its invariance under edge subdivision, making the kernels intrinsically well-suited for comparing graphs that represent different underlying spaces. We develop efficient algorithms for computing these kernels and analyze their complexity, showing that it depends primarily on the genus of the input graphs. Empirically, our kernels outperform existing methods in label-free settings, as demonstrated on both synthetic and real-world benchmark datasets. We further highlight their practical utility through an urban road network classification task.

The polysemantic nature of synthetic neurons in artificial intelligence language models is currently understood as the result of a necessary superposition of distributed features within the latent space. We propose an alternative approach, geometrically defining a neuron in layer n as a categorical vector space with a non-orthogonal basis, composed of categorical sub-dimensions extracted from preceding neurons in layer n-1. This categorical vector space is structured by the activation space of each neuron and enables, via an intra-neuronal attention process, the identification and utilization of a critical categorical zone for the efficiency of the language model - more homogeneous and located at the intersection of these different categorical sub-dimensions.

The Lovász hinge is a convex loss function proposed for binary structured classification, in which k related binary predictions jointly evaluated by a submodular function. Despite its prevalence in image segmentation and related tasks, the consistency of the Lovász hinge has remained open. We show that the Lovász hinge is inconsistent with its desired target unless the set function used for evaluation is modular. Leveraging the embedding framework of Finocchiaro et al. (2024), we find the target loss for which the Lovász hinge is consistent. This target, which we call the structured abstain problem, is a variant of selective classification for structured prediction that allows one to abstain on any subset of the k binary predictions. We derive a family of link functions, each of which is simultaneously consistent for all polymatroids, a subset of submodular set functions. We then give sufficient conditions on the polymatroid for the structured abstain problem to be tightly embedded by the Lovász hinge, meaning no target prediction is redundant. We experimentally demonstrate the potential of the structured abstain problem for interpretability in structured classification tasks. Finally, for the multiclass setting, we show that one can combine the binary encoding construction of Ramaswamy et al. (2018) with our link construction to achieve an efficient consistent surrogate for a natural multiclass generalization of the structured abstain problem.

While Knowledge Graphs (KGs) have become increasingly popular across various scientific disciplines for their ability to model and interlink huge quantities of data, essentially all real-world KGs are known to be incomplete. As such, with the growth of KG use has been a concurrent development of machine learning tools designed to predict missing information in KGs, which is referred to as the Link Prediction Task. The majority of state-of-the-art link predictors to date have followed an embedding-based paradigm. In this paradigm, it is assumed that the information content of a KG is best represented by the (individual) vector representations of its nodes and edges, and that therefore node and edge embeddings are particularly well-suited to performing link prediction. This thesis proposes an alternative perspective on the field's approach to link prediction and KG data modelling. Specifically, this work re-analyses KGs and state-of-the-art link predictors from a graph-structure-first perspective that models the information content of a KG in terms of whole triples, rather than individual nodes and edges. Following a literature review and two core sets of experiments, this thesis concludes that a structure-first perspective on KGs and link prediction is both viable and useful for understanding KG learning and for enabling cross-KG transfer learning for the link prediction task. This observation is used to create and propose the Structural Alignment Hypothesis, which postulates that link prediction can be understood and modelled as a structural task. All code and data used for this thesis are open-sourced. This thesis was written bilingually, with the main document in English and an informal extended summary in Irish. An Irish-language translation dictionary of machine learning terms (the Foclóir Tráchtais) created for this work is open-sourced as well.

We establish that testing for the equality of two probability measures on a general separable and compact metric space is equivalent to testing for the singularity between two corresponding Gaussian measures on a suitable Reproducing Kernel Hilbert Space. The corresponding Gaussians are defined via the notion of kernel mean and covariance embedding of a probability measure. Discerning two singular Gaussians is fundamentally simpler from an information-theoretic perspective than non-parametric two-sample testing, particularly in high-dimensional settings. Our proof leverages the Feldman-Hajek criterion for singularity/equivalence of Gaussians on Hilbert spaces, and shows that discrepancies between distributions are heavily magnified through their corresponding Gaussian embeddings: at a population level, distinct probability measures lead to essentially separated Gaussian embeddings. This appears to be a new instance of the blessing of dimensionality that can be harnessed for the design of efficient inference tools in great generality.

We present a dynamic self-balancing octree data structure that enables efficient neighborhood maintenance in evolving metric spaces, a key challenge in modern machine learning systems. Many learning and generative models operate as dynamical systems whose representations evolve during training, requiring fast, adaptive spatial organization. Our two-parameter octree supports logarithmic-time updates and queries, eliminating the need for costly full rebuilds as data distributions shift. We demonstrate its effectiveness in four areas: (1) accelerating Stein variational gradient descent by supporting more particles with lower overhead; (2) enabling real-time, incremental KNN classification with logarithmic complexity; (3) facilitating efficient, dynamic indexing and retrieval for retrieval-augmented generation; and (4) improving sample efficiency by jointly optimizing input and latent spaces. Across all applications, our approach yields exponential speedups while preserving accuracy, particularly in high-dimensional spaces where maintaining adaptive spatial structure is critical.

There is growing interest in developing statistical estimators that achieve exponential concentration around a population target even when the data distribution has heavier than exponential tails. More recent activity has focused on extending such ideas beyond Euclidean spaces to Hilbert spaces and Riemannian manifolds. In this work, we show that such exponential concentration in presence of heavy tails can be achieved over a broader class of parameter spaces called CAT($\kappa$) spaces, a very general metric space equipped with the minimal essential geometric structure for our purpose, while being sufficiently broad to encompass most typical examples encountered in statistics and machine learning. The key technique is to develop and exploit a general concentration bound for the Fr\'echet median in CAT($\kappa$) spaces. We illustrate our theory through a number of examples, and provide empirical support through simulation studies.

A Swiss and Austrian climbing pair have shattered the speed record for completing the daunting north faces of a famed trio of Swiss mountains - the Eiger, Mönch and Jungfrau. Switzerland's Nicolas Hojac and Austria's Philipp Brugger shaved nearly ten hours off the previous record set more than two decades ago.

The purpose of this paper is twofold. On a technical side, we propose an extension of the Hausdorff distance from metric spaces to spaces equipped with asymmetric distance measures. Specifically, we focus on the family of Bregman divergences, which includes the popular Kullback--Leibler divergence (also known as relative entropy). As a proof of concept, we use the resulting Bregman--Hausdorff divergence to compare two collections of probabilistic predictions produced by different machine learning models trained using the relative entropy loss. The algorithms we propose are surprisingly efficient even for large inputs with hundreds of dimensions. In addition to the introduction of this technical concept, we provide a survey. It outlines the basics of Bregman geometry, as well as computational geometry algorithms. We focus on algorithms that are compatible with this geometry and are relevant for machine learning.

Existing approaches remain largely constrained by traditional distance metrics, limiting their effectiveness in handling random data. In this work, we introduce the first k-means variant in the literature that operates within a probabilistic metric space, replacing conventional distance measures with a well-defined distance distribution function. This pioneering approach enables more flexible and robust clustering in both deterministic and random datasets, establishing a new foundation for clustering in stochastic environments. By adopting a probabilistic perspective, our method not only introduces a fresh paradigm but also establishes a rigorous theoretical framework that is expected to serve as a key reference for future clustering research involving random data. Extensive experiments on diverse real and synthetic datasets assess our model's effectiveness using widely recognized evaluation metrics, including Silhouette, Davies-Bouldin, Calinski Harabasz, the adjusted Rand index, and distortion. Comparative analyses against established methods such as k-means++, fuzzy c-means, and kernel probabilistic k-means demonstrate the superior performance of our proposed random normed k-means (RNKM) algorithm. Notably, RNKM exhibits a remarkable ability to identify nonlinearly separable structures, making it highly effective in complex clustering scenarios. These findings position RNKM as a groundbreaking advancement in clustering research, offering a powerful alternative to traditional techniques while addressing a long-standing gap in the literature. By bridging probabilistic metrics with clustering, this study provides a foundational reference for future developments and opens new avenues for advanced data analysis in dynamic, data-driven applications.