New methods for assessing and visualizing feature groups in machine learning models.
problem Lack of methods for interpreting feature groups in machine learning models.
method Permutation-based, refitting, and Shapley-based techniques for grouped feature importance. Introduced a sequential procedure for identifying stable feature combinations. Developed a combined features effect plot.
result Effective methods for assessing and visualizing the importance and effect of feature groups in machine learning models.
Proposes a differentiable hypergeometric distribution for learning group importance.
problem Learning the sizes of subsets in applications like clustering and weakly-supervised learning.
method Introduces a reparameterizable hypergeometric distribution to model group sizes and learn their relative importance.
result Outperforms previous methods in weakly-supervised learning and clustering.
Groups of importance in group theory have flexible stability properties.
problem Stability and flexibility of groups in geometric and combinatorial group theory.
method Establishing Kirchberg's Local Lifting Property and Lubotzky--Shalom's Property FD for specific groups.
result Groups like 3-manifold groups, limit groups, and certain one-relator groups are very flexibly stable. Improves model calibration and selection in unsupervised domain adaptation.
problem Distribution shifts in unsupervised domain adaptation.
method Developed a novel importance weighted group accuracy estimator.
result Improves state-of-the-art performances by 22% in model calibration and 14% in model selection.
In many application domains, it is important to characterize how complex learned models make their decisions across the distribution of instances. One way to do this is to identify the features and interactions among them that contribute to a model's predictive accuracy. We present a model-agnostic approach to this tas…
Survey on Coxeter groups for Lie group examples.
problem Understanding Coxeter groups and their applications.
method Constructing discrete subgroups of Lie groups using Coxeter groups.
result Coxeter groups provide new examples in discrete subgroups of Lie groups.
Group Shapley evaluates feature groups in business data, improving explainability in AI.
problem Evaluating the importance of feature groups in business and economic data.
method Developed Group Shapley and a significance testing procedure based on chi-square approximation.
result Market-related variables are identified as the most influential feature group.
Improves overparameterized models' robustness to distribution shifts.
problem Accuracy drop on testing distributions different from training.
method Importance tempering to improve decision boundaries.
result State-of-the-art results on worst group classification tasks.
AGS-CL selectively updates penalties based on node importance for continual learning.
problem Catastrophic forgetting in continual learning.
method Adaptive Group Sparsity (AGS) with proximal gradient descent.
result Significantly outperforms baselines on various continual learning benchmarks.
Study finds specific Lie groups with Kenmotsu structures.
problem Characterizing Lie groups with Kenmotsu structures.
method Determined Lie groups with left invariant Kenmotsu structures.
result These Lie groups are Einstein Riemannian manifolds.
New method for group representation presentations.
problem Constructing presentations for group representations.
method Developed a new technique for generating relations.
result Important for calculating Torelli group's homology.
We prove an equivariant version of the fact that word-hyperbolic groups have finite asymptotic dimension. This is important in connection with our forthcoming proof of the Farrell-Jones conjecture in algebraic K-theory for every word-hyperbolic group G and every coefficient ring R.
A novel validation method improves feature importance analysis in subject-specific ML models.
problem Limited effectiveness of general ML models trained on large datasets for individual patient outcomes.
method A novel validation approach using a general ML model with repeated trials and random seed variation.
result Consistent identification of key features at the subject level and improved group-level feature importance analysis.
Researchers find a way to bound the complexity of certain subgroup geometric invariants.
problem Understanding the geometric invariants of subgroups of direct products of free groups.
method Generalizing techniques for 'pushing fillings' into normal subgroups.
result Finitely presented subgroups of direct products of three free groups and subgroups of finiteness type Fn−1 in a direct product of n free groups have Dehn functions bounded above by N9. For groups of a topological origin, such as braid groups and mapping class groups, an important source of interesting and highly non-trivial representations is given by their actions on the twisted homology of associated spaces; these are known as homological representations. Representations of this kind have proved th…
In this easy introduction to higher gauge theory, we describe parallel transport for particles and strings in terms of 2-connections on 2-bundles. Just as ordinary gauge theory involves a gauge group, this generalization involves a gauge '2-group'. We focus on 6 examples. First, every abelian Lie group gives a Lie 2-gr…
In this note we find the metric of 6-dimensional h-space of the [33] type and then determine an important projective group characteristic of this h-space.
We give a combinatorial criterion that implies both the non-strong relative hyperbolicity and the one-endedness of a finitely generated group. We use this to show that many important classes of groups do not admit a strong relatively hyperbolic group structure and have one end. Applications include surface mapping clas…
In this paper we address the question of the existence of a model for the string 2-group as a strict Lie-2-group using the free loop group LSpin (or more generally LG for compact simple simply-connected Lie groups G). Baez-Crans-Stevenson-Schreiber constructed a model for the string 2-group using a based loop gro…
The 2-rank of a compact Lie group G is the maximal possible rank of the elementary 2-subgroup Z2×...Z2 of G. The study of 2-ranks (and p-rank for any prime p) of compact Lie groups was initiated in 1953 by A. Borel and J.-P. Serre. Since then the 2-ranks of compact Lie groups h…
We find a set of generators for the automorphism group of a graph product of finitely generated abelian groups entirely from a certain labeled graph. In addition, we find generators for the important subgroup of star-automorphisms defined in [7]. We follow closely the plan of M. Laurence's paper [11].
Two extremal classes of acyclic groups are discussed. For an arbitrary group G, there is always a homomorphism from an acyclic group of cohomological dimension 2 onto the maximum perfect subgroup of G, and there is always an embedding of G in a binate (hence acyclic) group. In the other direction, there are no nontrivi…
Overparameterized neural networks can be highly accurate on average on an i.i.d. test set yet consistently fail on atypical groups of the data (e.g., by learning spurious correlations that hold on average but not in such groups). Distributionally robust optimization (DRO) allows us to learn models that instead minimize…
We associate to a CAT(0)-space a flow space that can be used as the replacement for the geodesic flow on the sphere tangent bundle of a Riemannian manifold. We use this flow space to prove that CAT(0)-group are transfer reducible over the family of virtually cyclic groups. This result is an important ingredient in our …
Cone structures in quantum field theory linked to information geometry.
problem Understanding geometric structures in quantum field theory.
method Analyzing invariant cones under modular automorphism groups and their relation to Wishart laws.
result Explicit connection between CAH cones and Wishart laws.
The article compares predictor importance in classification problems with categorical outcomes.
problem Comparing predictor importance in classification problems with categorical response variables.
method The approach is based on the categorical Gini correlation (CGC) and tests differences in CGCs across predictor groups.
result The proposed methodology accommodates predictors of arbitrary and unequal dimensions and allows for dependence between predictor groups.
Pseudo-automorphisms are birational transformations acting as regular automorphisms in codimension 1. We import ideas from geometric group theory to prove that a group of birational transformations that satisfies a fixed point property on CAT(0) cubical complexes, for example a discrete countable group with Kazhdan Pro…
Veech groups are discrete subgroups of SL(2, R) which play an important role in the theory of translation surfaces. For a special class of translation surfaces called origamis or square-tiled surfaces their Veech groups are subgroups of finite index of SL(2, Z). We show that each stratum of the space of translation sur…
Study homology groups of mapping and Torelli groups for surfaces with abelian covers.
problem Understanding the homology of mapping and Torelli groups for surfaces with specific topologies.
method Examined the first homology group of mapping and Torelli groups with coefficients in the first rational homology group of the universal abelian cover of the surface.
result For surfaces with one boundary component, the twisted homology groups are finite-dimensional, but for surfaces with one puncture, they are infinite-dimensional.
In [V.O. Manturov, Non-reidemeister knot theory and its applications in dynamical systems, geometry, and topology, arxiv:1501.05208] the first named author gave the definition of k-free braid groups Gnk. Here we establish connections between free braid groups, classical braid groups and free groups: we describe e…
Homotopy classification for certain 4-manifolds with dihedral fundamental groups.
problem Classifying the homotopy types of specific 4-manifolds with dihedral fundamental groups.
method Using quadratic 2-type and combining with results from Hambleton-Kreck and Bauer.
result Homotopy types of finite oriented Poincaré 4-complexes are determined by their quadratic 2-type when fundamental group is dihedral.
A Lie group G endowed with a left invariant Riemannian metric g is called Riemannian Lie group. Harmonic and biharmonic maps between Riemannian manifolds is an important area of investigation. In this paper, we study different aspects of harmonic and biharmonic homomorphisms between Riemannian Lie groups. We show t…
We develop an analogue of the Birman exact sequence for the Torelli subgroup of Aut(F_n). This builds on earlier work of the authors who studied an analogue of the Birman exact sequence for the entire group Aut(F_n). These results play an important role in the authors' recent work on the second homology group of the To…
Interactive art piece shows braid groups and plane motions.
problem Understanding braid groups and their actions.
method Interactive art piece illustrating two braid group definitions and their action on free group.
result Demonstrates all plane motions in a single subspace.
We present a new method for manufacturing complex-valued harmonic morphisms from a wide class of Riemannian Lie groups. This yields new solutions from an important family of homogeneous Hadamard manifolds. We also give a new method for constructing left-invariant foliations on a large class of Lie groups producing harm…
We study isometric actions of finitely presented groups on R-trees. In this paper, we develop a relative version of the Rips machine to study pairs of such actions. An important example of a pair is a group action on an R-tree and a subgroup action on its minimal invariant su…
In 1964, John Stallings established an important relationship between the low-dimensional homology of a group and its lower central series. We establish a similar relationship between the low-dimensional homology of a group and its derived series. We also define a torsion-free-solvable completion of a group that is ana…
Adaptive Group Lasso selects important features in neural networks.
problem Lack of interpretability in neural networks.
method Adaptive Group Lasso for feature selection.
result Consistent feature selection for neural networks with theoretical guarantee.
Residual finiteness is known to be an important property of groups appearing in combinatorial group theory and low dimensional topology. In a recent work [2] residual finiteness of quandles was introduced, and it was proved that free quandles and knot quandles are residually finite. In this paper, we extend these resul…
We consider a homological enlargement of the mapping class group, defined by homology cylinders over a closed oriented surface (up to homology cobordism). These are important model objects in the recent Goussarov-Habiro theory of finite-type invariants of 3-manifolds. We study the structure of this group from several d…
Leveraging the intrinsic symmetries in data for clear and efficient analysis is an important theme in signal processing and other data-driven sciences. A basic example of this is the ubiquity of the discrete Fourier transform which arises from translational symmetry (i.e. time-delay/phase-shift). Particularly important…
Braid groups are an important and flexible tool used in several areas of science, such as Knot Theory (Alexander's theorem), Mathematical Physics (Yang-Baxter's equation) and Algebraic Geometry (monodromy invariants). In this note we will focus on their algebraic-geometric aspects, explaining how the representation the…
Fold maps are higher dimensional versions of Morse functions and fundamental and important tools in studying algebraic and differential topological properties of manifolds: as the theory established by Morse and the higher dimensional version, started by Thom and Whitney, later actively studied by Eliashberg, Levine et…
We give a brief overview of the current state of the study of the deformation theory of Kleinian groups. The topics covered include the definition of the deformation space of a Kleinian group and of several important subspaces; a discussion of the parametrization by topological data of the components of the closure of …
We investigate orthogonal representations of compact Lie groups from the point of view of their quotient spaces, considered as metric spaces. We study metric spaces which are simultaneously quotients of different representations and investigate properties of the corresponding representations. We obtain some structural …
Develops fair feature importance scores for tree-based models to interpret fairness.
problem Ensuring fairness in machine learning models, especially tree-based ones.
method Inspired by decision trees, proposes a novel fair feature importance score based on mean decrease in group bias.
result Valid interpretations of fairness for tree-based ensembles and surrogates of other ML systems.
Unified study of homological representations of mapping class groups.
problem Fundamental relationships between different homological representations.
method Analysis of twisted homology of configuration spaces.
result Unified account of non-degenerate pairings, embeddings, and isomorphisms.
Let Σg be a closed hyperbolic surface of genus g and let Ham(Σg) be the group of Hamiltonian diffeomorphisms of Σg. The most natural word metric on this group is the autonomous metric. It has many interesting properties, most important of which is the bi-invariance of this metric. In this work we show that $…