Research
On-device research index

arXiv research

A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.

169,181 papers · 148 categories

Trend · papers per month

2855708551,140 · Jun 202019922001200920182026
48 results for Classification performance

Few-shot image classification is improved by correcting CNNs' texture bias.

problem Few-shot image classification performance is hindered by CNNs' texture bias.
method Corrected CNNs' texture bias using a simpler method than state-of-the-art approaches.
result State-of-the-art performance on miniImageNet task achieved.

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.

Paper introduces LPCs for robust classification with performance bounds.

problem Conventional classification techniques constrain rules and use surrogate losses.
method Robust risk minimization (RRM) for unconstrained classification rules, optimizing 0-1 loss.
result LPCs provide performance bounds and competitive performance with state-of-the-art techniques.

Enhances classification performance with small, additive perturbations.

problem Improving classification performance using small, additive perturbations.
method Proposes a perturbation generation network (PGN) based on adversarial learning to enhance classifier performance.
result Demonstrates that PGN can enhance overall classification performance without altering the target classifier network.

Study finds resolution impacts human classification performance in MNIST data.

problem Understanding factors affecting human classification performance in machine learning.
method Empirical study of MNIST data at various resolutions.
result Derived a quantitative relationship between resolution and human classification performance.

The paper critiques current time series classification evaluation methods.

problem Current performance evaluation methods in time series classification are criticized.
method No specific new method proposed, but a critical analysis of existing methods.
result Suggests a need for discussion and reflection on TSC performance evaluation.

Optimizes SGD for text classification using hyper-parameter tuning.

problem Improving performance of SGD in text classification.
method Fine-tuning hyper-parameters using Grid-Search approach.
result Optimized SGD classification leads to better accuracy and execution time.

Temporal information impacts only a fraction of time series datasets, skewing benchmark evaluations.

problem Temporal information's impact on time series classification is often overestimated.
method Permutation tests on UCR archive to identify datasets where temporal info is irrelevant.
result Many tabular datasets perform well without temporal info, skewing benchmark evaluations.

Study identifies pitfalls in assessing hierarchies for multi-class classification.

problem Lack of understanding in selecting hierarchies for multi-class classification.
method Analyzed and compared popular approaches to extracting hierarchies.
result Hierarchy quality becomes irrelevant when using powerful classifiers.

C-HMCNN(h) improves HMC classification by leveraging class hierarchy.

problem Hierarchical multi-label classification with class hierarchy constraints.
method Exploits class hierarchy to produce coherent predictions for multi-label classification.
result C-HMCNN(h) outperforms state-of-the-art models in HMC classification.

Paper develops new methods for binary classification with complex performance measures.

problem Complex performance measures in binary classification are not decomposable and require new theoretical and methodological developments.
method Identifies Karmic and threshold-quasi-concavity properties, and develops a computationally practical plug-in classifier.
result Bayes optimal classifier is a threshold function of conditional probability, leading to practical classification error analysis.

Study proposes an ensemble learning method to improve multi-label classification performance.

problem Improving multi-label classification performance in machine learning.
method Ensemble learning approach using multiple base-level algorithms.
result Proposed method outperforms base-level algorithms in multi-label classification.

A deep learning approach classifies medical images hierarchically.

problem Limitations of traditional supervised classifiers in medical image classification.
method Hierarchical Medical Image Classification (HMIC) using deep learning models.
result HMIC achieved better performance in classifying medical images hierarchically.

A method selects key genes from tumor transcriptomics data using kernel methods and improves classification performance.

problem Feature selection for tumor classification using gene expression data.
method Multiple Kernel Learning with latent regularization and non-linear dimensionality reduction.
result Improved tumor classification performance on unseen test samples.

New framework models algorithmic decisions affecting data distribution, enabling efficient learning.

problem Algorithmic decisions can alter data distribution, affecting model performance.
method Model performative effects as push-forward measures, estimating gradients under shift operators.
result Prove convexity of performative risk, allowing more accurate models to be harder to classify.

End-to-end neural network extracts graph structure from EEG signals for improved emotional video classification.

problem Challenges in achieving accurate EEG classification for emotional video analysis.
method Proposes an end-to-end neural network model that learns an appropriate multi-layer graph structure from raw EEG signals.
result Improves performance in emotional video classification compared to manually defined connectivity structures.

DS techniques outperform K-NN in classification accuracy.

problem Improving classification performance using multiple classifier systems.
method Dynamic Selection (DS) compared to K-NN, focusing on neighborhood quality and instance hardness.
result DS techniques achieve higher classification accuracy than K-NN.

Ladder Networks improve semi-supervised hyperspectral image classification.

problem Semi-supervised hyperspectral image classification with limited labeled data.
method Jointly optimizing a supervised and unsupervised cost in a Ladder Network.
result Convolutional Ladder Network achieves state-of-the-art performance with minimal labeled data.

Machine learning classification limits estimated using Kullback-Leibler divergence and Cohen's Kappa.

problem Estimating the best possible performance of machine learning classification algorithms.
method Relating Kullback-Leibler divergence to Cohen's Kappa and using the Chernoff-Stein Lemma to estimate error rates.
result Classification algorithms could not have performed any better due to underlying probability density functions for the two classes.

Paper discovers shifting patterns in sequence classification and proposes a method to improve performance.

problem Discriminative patterns in sequential data are time-dependent and degrade traditional classification methods.
method Proposes a novel sequence classification method using multi-instance learning and LSTM models to detect and model shifting patterns.
result Demonstrates superior sequence classification performance and detection of shifting patterns in cropland mapping and affective state recognition.

Research aims to explain how ResNets' stability improves image classification performance.

problem Understanding why ResNets enhance image classification performance.
method Examines batch normalization and the dynamical systems view of ResNets to understand stability and smoothness.
result Stability of inter-layer propagation in ResNets contributes to enhanced performance.

Study improves pollen detection in optical and holographic images using deep learning.

problem Improving pollen detection accuracy in holographic microscopy images.
method Used YOLOv8s for detection and MobileNetV3L for classification, addressing performance gaps through dataset expansion and automated labeling.
result Significant improvement in detection and classification performance on holographic images.

HexaGAN tackles real-world classification issues with missing data, class imbalance, and missing labels.

problem Missing data, class imbalance, and missing labels in real-world data.
method Generative adversarial network framework with six components and novel loss functions.
result Up to 5% improvement in classification performance compared to state-of-the-art methods.

This work examines how non-identical data distributions affect Federated Learning performance.

problem The impact of non-identical data distributions on Federated Learning performance.
method Synthesized datasets with varying degrees of data distribution similarity, evaluated Federated Averaging algorithm performance, proposed server momentum mitigation.
result Performance of Federated Learning degrades as data distributions differ more, and a mitigation strategy improves accuracy.

Paper shows similarity learning can lead to strong binary classification performance.

problem How similarity learning can lead to good classification performance.
method Product-type formulation of similarity learning is connected to binary classification through an excess risk bound.
result Similarity learning can directly elicit a decision boundary for binary classification.

The study analyzes performance indices for class-imbalanced data and identifies conditions they must meet.

problem Distortions in performance indices under class imbalance.
method Identified two conditions for performance indices and analyzed four binary and five multi-class indices.
result Recommended appropriate indices for evaluating classifiers in class-imbalanced scenarios.

Unified framework for classification and latent variable extraction.

problem Combining supervised and unsupervised learning for robust classification.
method Hierarchical Bayesian model divided into three stages.
result Improved classification performance in hyperspectral image interpretation.

Paper shows adversarial classification algorithms are inherently more sensitive to data manipulation.

problem Inadequate provable guarantees for machine learning performance, especially in unreliable data environments.
method Formal analysis of binary classification algorithms' sensitivity to adversarial manipulation.
result Fundamental tradeoff curve between accuracy and sensitivity is determined by data statistics, not algorithm tuning.

Paper analyzes cyber risk classifications for forecasting performance.

problem Lack of effective out-of-sample forecasting performance in current cyber risk classifications.
method Rolling window analysis using threshold weighted scoring functions.
result Dynamic and impact-based cyber risk classifiers outperform others in forecasting future cyber risk losses.

Novel IndRNN model improves seizure/non-seizure classification accuracy.

problem Manual EEG analysis by neurologists is laborious and prone to errors.
method Leverages independently recurrent neural networks (IndRNN) to capture seizure features across various time scales.
result The proposed approach outperforms state-of-the-art methods in cross-validation experiments.

A new classification rule for FDA improves classification performance by accounting for unequal covariance matrices.

problem Unequal covariance matrices in practical situations affect the performance of FDA and its variants.
method Proposes a novel classification rule for FDA that accounts for unequal covariance matrices, applicable to many FDA variants.
result The new classification rule improves classification performance compared to original FDA and variants.

This study compares different data preprocessing methods for gait analysis.

problem Improving the generalizability of machine learning models in gait analysis.
method Compared various data preprocessing steps including filtering, time derivative, normalization, and scaling.
result Different preprocessing combinations affect gait classification performance.