Interactive tool helps choose and understand classification metrics.
problem Common metrics for binary classification have limitations.
method Graphical application to visualize and explore evaluation metrics.
result Promotes careful attention to interpretation of metrics.
Unified framework for comparing classification metrics across different imbalance rates.
problem Differences in scale and sensitivity to class imbalance rates in classification metrics.
method Introduces outperformance standardization (OPS) function to map metrics to a common scale.
result Unified o-value metric provides clear comparison across different imbalance rates.
This paper reviews metrics for evaluating multi-class classification models.
problem Evaluating and comparing multi-class classification models.
method Review and analysis of metrics.
result Promising multi-class metrics are highlighted and their usages are demonstrated.
A new robust time series distance metric for k-NN classification.
problem Robustness against arbitrary data contamination in time series classification.
method Proposes a novel distance metric with worst-case O(nlogn) complexity. result Demonstrates competitive classification accuracy in k-NN time series classification.
Classifies Kähler metrics with constant holomorphic curvature.
problem Classifying Kähler metrics with constant holomorphic sectional curvature.
method Exploiting the geometry of the bundle of 1-jets of holomorphic functions.
result Local classification of Kähler metrics with constant holomorphic sectional curvature.
Enhanced metrics for multiclass classification improve on existing methods.
problem Lack of decisive poor classification results in existing multiclass metrics.
method Introduces three new metrics derived from multivariate Pearson correlation coefficients.
result New metrics decisively indicate poor classification results.
Proposes sigmoidF1 loss for multilabel classification, improving performance metrics.
problem Lack of smooth, tractable loss functions for multilabel classification.
method Introduces sigmoidF1, a smooth F1 score surrogate loss function.
result sigmoidF1 outperforms other loss functions on various datasets and metrics.
Completed classification of Einstein spaces with specific metric properties.
problem Classifying Einstein spaces with a specific type of Stackel metric.
method Invariant under three-parameter abelian group of motions, completed classification of vacuum and electrovacuum spaces.
result Complete list of metrics for Einstein spaces in privileged coordinate systems.
This thesis formalizes metric selection for machine learning applications.
problem Selecting the right performance metric for machine learning models.
method Metric Elicitation framework using pairwise preference feedback.
result Novel strategies for eliciting various classification metrics robust to noise.
The paper classifies 3D paracontact and almost paracosymplectic spaces.
problem Classifying 3D paracontact and almost paracosymplectic spaces.
method Detailed structure analysis and local classification for all possible values of κ.
result Local classification of paracontact metric and almost paracosymplectic (κ,μ)-spaces for every possible value of κ.
Study classifies Kähler-Einstein metrics with rotational symmetries.
problem Classify Kähler-Einstein metrics with rotational symmetries.
method Focus on integrable structures to classify metrics.
result Classify Kähler-Einstein metrics with rotational symmetries.
The paper classifies spherically symmetric Finsler metrics as Berwaldian or Riemannian.
problem Classifying spherically symmetric Finsler metrics.
method Proving all spherically symmetric Landsberg surfaces are Berwaldian and modifying the classification of spherically symmetric Finsler metrics.
result All Berwald spherically symmetric metrics of dimension n≥3 are Riemannian or given by a specific formula. New metrics improve performance in imbalanced classification problems.
problem Established metrics favor classifiers ignoring minority classes.
method Introduce robust modifications of F-score and MCC.
result TPR is bounded away from 0 in imbalanced settings.
Metric learning enhances combinatorial coverage metrics' ability to predict classification errors.
problem Dataset dependence of combinatorial coverage metrics in anticipating classification errors.
method Metric learning to improve latent space separation of data classes.
result Metric learning increases SDCCMs' ability to distinguish between correctly and incorrectly classified data.
Using the twistor correspondence, we give a classification of toric anti-self-dual Einstein metrics: each such metric is essentially determined by an odd holomorphic function. This explains how the Einstein metrics fit into the classification of general toric anti-self-dual metrics given in an earlier paper (math.DG/06…
This paper introduces a new metric for deep learning networks based on their classification performance.
problem The mystery and black-box nature of deep learning networks.
method Proposes a new distance measure based on the probabilistic performance of deep learning networks.
result The proposed metric space is compact and coincides with the quotient topological space.
We provide a general theoretical analysis of expected out-of-sample utility, also referred to as decision-theoretic classification, for non-decomposable binary classification metrics such as F-measure and Jaccard coefficient. Our key result is that the expected out-of-sample utility for many performance metrics is prov…
Clustering and classification critically rely on distance metrics that provide meaningful comparisons between data points. We present mixed-integer optimization approaches to find optimal distance metrics that generalize the Mahalanobis metric extensively studied in the literature. Additionally, we generalize and impro…
The paper classifies left-invariant pseudo-Riemannian metrics on specific Lie groups.
problem Classifying left-invariant pseudo-Riemannian metrics on Lie groups.
method Analyzing left-invariant metrics on specific Lie groups with n≥4. result A complete classification of left-invariant pseudo-Riemannian metrics for Lie groups of dimension n≥4. Study creates web interface to elicit user-preferred metrics.
problem Eliciting classification metrics that align with user preferences.
method Developed a web-based interface and conducted a user study.
result Users preferred metrics that align with their task and context.
Investigates hypersurfaces in Finsler spaces with generalized square metrics.
problem Classifying and understanding hypersurfaces in Finsler spaces with generalized square metrics.
method Examined the generalized square metric F(x,y) = (α(x,y) + β(x,y))^(n+1)/(α^n(x,y)) and its application to Finslerian hypersurfaces.
result Established the classification and existence of first, second, and third kind of hyperplanes in the Finsler manifold.
The paper classifies metrics on Heisenberg group's cotangent bundle.
problem Investigating moduli spaces of left invariant metrics on cotangent bundles of Heisenberg group.
method Algebraic approach combined with geometrical tools like classification of hyperbolic plane conics.
result Detailed classification of various types of metrics and their properties.
We complete a minor gap in Gromoll and Walschap classification of metric fibrations from the Euclidean space, thus completing the classification of Riemannian foliations on Euclidean spaces.
We investigate a certain class of solvable metric Lie algebras. For this purpose a theory of twofold extensions associated to an orthogonal representation of an abelian Lie algebra is developed. Among other things, we obtain a classification scheme for indecomposable metric Lie algebras with maximal isotropic centre an…
We study the quantification of uncertainty of Convolutional Neural Networks (CNNs) based on gradient metrics. Unlike the classical softmax entropy, such metrics gather information from all layers of the CNN. We show for the EMNIST digits data set that for several such metrics we achieve the same meta classification acc…
Classifies 4D toric Hermitian ALF metrics with conical singularities.
problem Classifying specific types of 4D Riemannian metrics.
method Explicit formulas provided for classification.
result Examples of metrics with conical singularities have infinitely many distinct topologies.
The paper classifies special types of contact metric manifolds with curvature conditions.
problem Classifying N(κ)-contact metric manifolds with specific curvature tensors. method Examining flatness conditions on T-curvature tensor and analyzing specific curvature tensors. result A classification of N(κ)-contact metric manifolds under various curvature conditions. In this paper, the concept of isotropic projective Ricci curvature has been investigated. By classification of Randers metric of isotropic projective Ricci curvature, it is shown that Randers metric of projective Ricci curvature is reversible if and only if it is of square projective Ricci curvature.
Paper develops a framework to optimize neural networks using weighted metrics.
problem Discrepancy between maximizing weighted classification scores and minimizing loss function.
method Formalizes weighted classification metrics and constructs corresponding losses.
result Framework includes well-established approaches like cost-sensitive learning and weighted cross entropy.
We study the construction and classification of weakly Bochner-flat (WBF) metrics (i.e., Kahler metrics with coclosed Bochner tensor) on compact complex manifolds. A Kahler metric is WBF if and only if its `normalized' Ricci form is a hamiltonian 2-form: such 2-forms were introduced and studied in previous papers in th…
The author is planning if possible classify all three-dimensional (κ,μ)-manifolds wether contact metric, almost cosymplectic, para-contact metric, almost para-cosymplectic. Of course classification in contact or almost cosymplectic cases already is provdied. Up to authors knowledge there is no classification for para…
This work improves metric learning models by incorporating class hierarchies.
problem Class hierarchies are often ignored in classification-based metric learning models.
method Trained softmax classifier and metric learning models with predefined class hierarchies.
result ProxyDR model performs better in hierarchical inference and hierarchy-informed performance.
Traditional text classifiers are limited to predicting over a fixed set of labels. However, in many real-world applications the label set is frequently changing. For example, in intent classification, new intents may be added over time while others are removed. We propose to address the problem of dynamic text classifi…
Paper tackles imbalanced binary classification by optimizing precision and recall directly.
problem Imbalanced binary classification where standard accuracy is misleading.
method Exact constrained reformulations for precision and recall optimization.
result ERO framework outperforms state-of-the-art methods on multiple datasets.
Determining the associated metrics we get a local classification of contact metric three manifolds.
New method accelerates large margin metric learning for nearest neighbor classification.
problem Efficiently learning metrics for nearest neighbor classification.
method Triplet mining and stratified sampling for large margin metric learning.
result Improved efficiency and scalability of optimization.
Proposes an online metric learning method for multi-label classification.
problem Lack of consideration for label dependencies and theoretical analysis of loss functions in existing multi-label classification methods.
method Develops a novel online metric learning paradigm based on k-Nearest Neighbour (kNN) and large margin principle, adapted for online streaming data.
result The proposed OML algorithm outperforms state-of-the-art methods on benchmark multi-label datasets.
Recent work in metric learning has significantly improved the state-of-the-art in k-nearest neighbor classification. Support vector machines (SVM), particularly with RBF kernels, are amongst the most popular classification algorithms that uses distance metrics to compare examples. This paper provides an empirical analy…
New algorithms optimize metrics for binary classification with class imbalance.
problem Optimizing metrics like Fβ, AM, Jaccard for imbalanced classes.
method Reformulates metric optimization as cost-sensitive learning, using surrogate loss functions.
result METRO algorithms provide strong theoretical guarantees and outperform baselines.
This paper classifies hypersurfaces in n+1 with parallel Fubini-Pick form.
problem Classifying hypersurfaces with parallel Fubini-Pick form in \(\mathbb{R}^{n+1}\).
method Defining a generalized Calabi product and proving decomposition theorems.
result Complete classification of Calabi hypersurfaces in \(\mathbb{R}^{n+1}\) with parallel Fubini-Pick form.
Proposes a method to select fair performance metrics through metric elicitation.
problem Choosing fair performance metrics in multiclass classification with multiple sensitive groups.
method Metric elicitation strategy that requires only relative preference feedback and is robust to noise.
result Elicits group-fair performance metrics for multiclass classification problems.
In this paper we classify all simply connected five dimensional nilpotent Lie groups which admit (α,β)-metrics of Berwald and Douglas type defined by a left invariant Riemannian metric and a left invariant vector field. During this classification we give the geodesic vectors, Levi-Civita connection, curvature tensor,…
Classifies homogeneous Riemannian structures on 3D Lie groups.
problem Classifying homogeneous Riemannian structures on 3D Lie groups.
method Classification based on left invariant metrics and previous classifications.
result Complete classification of homogeneous Riemannian structures on 3D Lie groups.
Completes the classification of Moebius deformable hypersurfaces for dimensions 5 and above.
problem Missing examples in the classification of Moebius deformable hypersurfaces for dimensions 5 and above.
method Investigates the class of Moebius deformable hypersurfaces and completes the classification for dimensions 5 and above.
result Completes the classification of Moebius deformable hypersurfaces for dimensions 5 and above.
In this paper, we study locally projectively flat Finsler metrics with constant flag curvature K. We prove those are totally determined by their behaviors at the origin by solving some nonlinear PDEs. The classifications when K=0, K=−1 and K=1 are given respectively in an algebraic way.…
This paper uses UOT metrics for better dimensionality reduction and classification/clustering.
problem Improving dimensionality reduction and classification/clustering methods.
method Uses Hellinger--Kantorovich metric from unbalanced optimal transport (UOT).
result UOT outperforms Euclidean and OT-based methods in classification and clustering tasks.
Classifies geodesic flows on projective plane with potential field.
problem Classifying geodesic flows on a projective plane with a potential field.
method Liouville classification and calculation of Fomenko--Zieschang invariants.
result All Fomenko--Zieschang invariants of the system are calculated.
We use Bott-Chern cohomology to measure the non-Kählerianity of 6-dimensional nilmanifolds endowed with the invariant complex structures in M. Ceballos, A. Otal, L. Ugarte, and R. Villacampa's classification, [Invariant Complex Structures on 6-Nilmanifolds: Classification, Frölicher Spectral Sequence and Special Hermit…