arXiv research

Paper explains distance-based classifiers using neural network structures.

We consider classifiers for high-dimensional data under the strongly spiked eigenvalue (SSE) model. We first show that high-dimensional data often have the SSE model. We consider a distance-based classifier using eigenstructures for the SSE model. We apply the noise reduction methodology to estimation of the eigenvalue…

Time series classification is an increasing research topic due to the vast amount of time series data that are being created over a wide variety of fields. The particularity of the data makes it a challenging task and different approaches have been taken, including the distance based approach. 1-NN has been a widely us…

The reliable measurement of confidence in classifiers' predictions is very important for many applications and is, therefore, an important part of classifier design. Yet, although deep learning has received tremendous attention in recent years, not much progress has been made in quantifying the prediction confidence of…

Improves classification of microbiome data using mixture distributions.

The study sets limits on how robust classifiers can be against adversarial attacks.

A statistical model predicts generalization in few-shot learning.

The Mahalanobis distance-based confidence score, a recently proposed anomaly detection method for pre-trained neural classifiers, achieves state-of-the-art performance on both out-of-distribution (OoD) and adversarial examples detection. This work analyzes why this method exhibits such strong performance in practical s…

Extends RF proximities to all supervised distance-based machine learning contexts.

ClusTR improves clustering-based models' robustness without adversarial training.

Despite numerous attempts to defend deep learning based image classifiers, they remain susceptible to the adversarial attacks. This paper proposes a technique to identify susceptible classes, those classes that are more easily subverted. To identify the susceptible classes we use distance-based measures and apply them …

Transforms distance-based outlier scores into interpretable probabilistic estimates.

Improves uncertainty estimation and OOD detection in neural networks.

Random forest can be adapted for open-set recognition with improved performance.

A scalable version of MADD improves big-data classification speed.

Deep neural networks (DNNs) are poorly calibrated when trained in conventional ways. To improve confidence calibration of DNNs, we propose a novel training method, distance-based learning from errors (DBLE). DBLE bases its confidence estimation on distances in the representation space. In DBLE, we first adapt prototypi…

Efficient method classifies locally stationary time series based on second-order characteristics.

Neural networks learn distance-based representations, not just intensity.

DSI measures dataset separability for neural networks.

PAC-Bayesian bounds improve understanding of K-NN classifier performance.

The paper improves uncertainty quantification for node classification using distance-based regularization.

Solves clustering contradictions by high-dimensional embedding with wide gaps.

BRDAD uses bagging and regularization to improve anomaly detection without labeled data.

Flexible classifier using Mahalanobis distances for non-elliptical distributions.

The $k$-nearest neighbour ($k$-NN) classifier is one of the oldest and most important supervised learning algorithms for classifying datasets. Traditionally the Euclidean norm is used as the distance for the $k$-NN classifier. In this thesis we investigate the use of alternative distances for the $k$-NN classifier. We …

Unified approach to multiclass classification using Gabriel graphs.

A new measure identifies clusters without assuming data distribution.

A method for ranking items using distance-based learning from positive and unlabeled data.

Anomalies (unusual patterns) in time-series data give essential, and often actionable information in critical situations. Examples can be found in such fields as healthcare, intrusion detection, finance, security and flight safety. In this paper we propose new conformalized density- and distance-based anomaly detection…

Due to the popularity of the Internet and smart mobile devices, more and more financial transactions and activities have been digitalized. Compared to traditional financial fraud detection strategies using credit-related features, customers are generating a large amount of unstructured behavioral data every second. In …

For time series comparisons, it has often been observed that z-score normalized Euclidean distances far outperform the unnormalized variant. In this paper we show that a z-score normalized, squared Euclidean Distance is, in fact, equal to a distance based on Pearson Correlation. This has profound impact on many distanc…

Neural networks can learn distance metrics affecting model performance.

GICDM corrects hubness in embedding spaces for better generative model evaluation.

PAC-Bayesian set up involves a stochastic classifier characterized by a posterior distribution on a classifier set, offers a high probability bound on its averaged true risk and is robust to the training sample used. For a given posterior, this bound captures the trade off between averaged empirical risk and KL-diverge…

A new Wasserstein $K$-means method for clustering probability distributions.

In general, the clustering problem is NP-hard, and global optimality cannot be established for non-trivial instances. For high-dimensional data, distance-based methods for clustering or classification face an additional difficulty, the unreliability of distances in very high-dimensional spaces. We propose a distance-ba…

Learning expressive low-dimensional representations of ultrahigh-dimensional data, e.g., data with thousands/millions of features, has been a major way to enable learning methods to address the curse of dimensionality. However, existing unsupervised representation learning methods mainly focus on preserving the data re…

In light of the power problems of statistical tests and undisciplined use of alpha-based statistics to compare models, this paper proposes a unified set of distance-based performance metrics, derived as the square root of the sum of squared alphas and squared standard errors. The Bayesian investor views model performan…

New metric learning approach for tree data reduces computation cost.

Clustering is an essential data mining tool that aims to discover inherent cluster structure in data. As such, the study of clusterability, which evaluates whether data possesses such structure, is an integral part of cluster analysis. Yet, despite their central role in the theory and application of clustering, current…

Discriminative neural networks address class imbalance in coronary heart disease risk analysis.

A new robust time series distance metric for k-NN classification.

Distance plays a fundamental role in measuring similarity between objects. Various visualization techniques and learning tasks in statistics and machine learning such as shape matching, classification, dimension reduction and clustering often rely on some distance or similarity measure. It is of tremendous importance t…

Distance-based tests, also called "energy statistics", are leading methods for two-sample and independence tests from the statistics community. Kernel-based tests, developed from "kernel mean embeddings", are leading methods for two-sample and independence tests from the machine learning community. A fixed-point transf…

New GP kernel handles mixed-categorical data, improving model accuracy.

Study evaluates initialization strategies for infinite hidden Markov models.

Outliers are ubiquitous in modern data sets. Distance-based techniques are a popular non-parametric approach to outlier detection as they require no prior assumptions on the data generating distribution and are simple to implement. Scaling these techniques to massive data sets without sacrificing accuracy is a challeng…

This paper explores ratio-based loss functions for machine learning.