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.

168,657 papers · 148 categories

Trend · papers per month

count % of corpus

7 results for “UTKFace”

FlowSDR: Sufficient Dimension Reduction via Conditional Normalizing Flowsstat.ME

FlowSDR learns a low-dimensional projection preserving the response's conditional distribution.

problem Learning a low-dimensional projection that captures the response's conditional distribution.

method FlowSDR uses conditional log-likelihood maximization with monotone rational-quadratic spline flows to learn the projection and conditional density.

result FlowSDR outperforms existing SDR methods in various simulation settings and a face-age prediction task.

Sufficient Dimension Reduction Conditional Normalizing Flows Mutual Information Face Age Prediction UTKFace

2026-05-31abs ↗pdf ↗

Conformal Risk Control for Ordinal Classificationcs.LG

Develops methods to control risk in ordinal classification tasks.

problem Controlling risk in ordinal classification tasks.

method Formulated ordinal classification in conformal risk control framework, proposed loss functions and algorithms.

result Demonstrated effectiveness and analyzed differences in risk control methods.

conformal prediction ordinal classification risk control loss functions diabetic retinopathy

2024-05-01abs ↗pdf ↗

The Promises and Pitfalls of Deep Kernel Learningstat.ML

DKL combines neural networks and Gaussian processes but can overfit.

problem Overfitting in DKL models.

method Careful experimentation on various datasets and investigation of optimization dynamics.

result Overfitting from DKL can be worse than non-Bayesian models.

Deep Kernel Learning Gaussian Processes Overfitting Uncertainty Estimation Optimization Dynamics

2021-02-24abs ↗pdf ↗

Non-linear Neurons with Human-like Apical Dendrite Activationscs.NE

In order to classify linearly non-separable data, neurons are typically organized into multi-layer neural networks that are equipped with at least one hidden layer. Inspired by some recent discoveries in neuroscience, we propose a new model of artificial neuron along with a novel activation function enabling the learni…

2020-02-02abs ↗pdf ↗

Distilling and Transferring Knowledge via cGAN-generated Samples for Image Classification and Regressioncs.CV

Unified framework for image classification and regression using cGAN-generated samples.

problem Lack of unified KD methods for both classification and regression tasks.

method cGAN-KD framework based on cGANs.

result Unified framework for both classification and regression tasks, compatible with other KD methods.

Knowledge distillation cGANs image classification image regression error bound

2021-04-07abs ↗pdf ↗

Which Shortcut Cues Will DNNs Choose? A Study from the Parameter-Space Perspectivecs.LG

Study reveals DNNs prefer easy-to-learn cues over essential ones in image recognition.

problem DNNs learn easy-to-learn features that aren't essential to the task.

method WCST-ML training setup with shortcut cues on synthetic and face datasets.

result DNNs converge to solutions focusing on preferred cues, leading to flat minima.

shortcut learning deep neural networks image recognition flat minima Kolmogorov complexity

2021-10-06abs ↗pdf ↗

Continuous Conditional Generative Adversarial Networks: Novel Empirical Losses and Label Input Mechanismscs.CV

CcGAN tackles conditional image generation for continuous labels.

problem Mathematical challenges in conditioning on continuous, scalar labels.

method Proposes novel empirical losses and label input methods for continuous conditional GANs.

result CcGAN generates diverse, high-quality images from continuous labels.

Conditional GAN Continuous Labels Image Generation Empirical Losses Label Input

2020-11-15abs ↗pdf ↗