arXiv research

Designs an online selective sampling approach for choosing which model to use.

The paper revisits discriminative vs. generative classifiers, showing naive Bayes requires fewer samples.

Improved robustness of BERT through orthogonal classifier heads.

A RL-based method adds conditional controls to pre-trained diffusion models.

This paper presents our contribution to PolEval 2019 Task 6: Hate speech and bullying detection. We describe three parallel approaches that we followed: fine-tuning a pre-trained ULMFiT model to our classification task, fine-tuning a pre-trained BERT model to our classification task, and using the TPOT library to find …

Study shows pre-trained models can handle long-tailed relations well, improving classifier performance.

Large-scale datasets may contain significant proportions of noisy (incorrect) class labels, and it is well-known that modern deep neural networks (DNNs) poorly generalize from such noisy training datasets. To mitigate the issue, we propose a novel inference method, termed Robust Generative classifier (RoG), applicable …

A new method aggregates generative classifiers to resist adversarial attacks.

Improved speech emotion recognition using pre-trained language models.

Adversarial examples in machine learning for images are widely publicized and explored. Illustrations of misclassifications caused by slightly perturbed inputs are abundant and commonly known (e.g., a picture of panda imperceptibly perturbed to fool the classifier into incorrectly labeling it as a gibbon). Similar atta…

Deep neural networks have been demonstrated to be vulnerable to adversarial attacks, where small perturbations intentionally added to the original inputs can fool the classifier. In this paper, we propose a defense method, Featurized Bidirectional Generative Adversarial Networks (FBGAN), to extract the semantic feature…

A test detects unfairness in machine learning classifiers.

Mahalanobis distance detects anomalies well, but not for classification.

Revises Bayesian model averaging for foundation models.

This paper analyzes kNN convergence over feature transformations.

Deep neural networks have been widely deployed in various machine learning tasks. However, recent works have demonstrated that they are vulnerable to adversarial examples: carefully crafted small perturbations to cause misclassification by the network. In this work, we propose a novel defense mechanism called Boundary …

Recent works have shown the effectiveness of randomized smoothing as a scalable technique for building neural network-based classifiers that are provably robust to $\ell_2$-norm adversarial perturbations. In this paper, we employ adversarial training to improve the performance of randomized smoothing. We design an adap…

In this paper, we present a statistical-mechanical analysis of deep learning. We elucidate some of the essential components of deep learning---pre-training by unsupervised learning and fine tuning by supervised learning. We formulate the extraction of features from the training data as a margin criterion in a high-dime…

In this study, a novel sparsity-driven weighted ensemble classifier (SDWEC) that improves classification accuracy and minimizes the number of classifiers is proposed. Using pre-trained classifiers, an ensemble in which base classifiers votes according to assigned weights is formed. These assigned weights directly affec…

Robust cancer screening model using pre-trained ensembles for biomarkers.

Two things seem to be indisputable in the contemporary deep learning discourse: 1. The categorical cross-entropy loss after softmax activation is the method of choice for classification. 2. Training a CNN classifier from scratch on small datasets does not work well. In contrast to this, we show that the cosine loss fun…

Adversarial perturbations dramatically decrease the accuracy of state-of-the-art image classifiers. In this paper, we propose and analyze a simple and computationally efficient defense strategy: inject random Gaussian noise, discretize each pixel, and then feed the result into any pre-trained classifier. Theoretically,…

S2OSC improves OSC by filtering and re-training models with out-of-class instances.

New classifiers ensure fairness by adjusting a base classifier's operating characteristics.

TR0N turns pre-trained models into conditional ones with minimal training.

Deep neural networks have shown promising results for various clinical prediction tasks such as diagnosis, mortality prediction, predicting duration of stay in hospital, etc. However, training deep networks -- such as those based on Recurrent Neural Networks (RNNs) -- requires large labeled data, high computational res…

This work investigates how neural collapse improves transfer learning for large-scale models.

Paper introduces active Bayesian method for assessing black-box classifiers efficiently.

GP-TS optimizes TLM pre-training hyperparameters efficiently.

We propose a robust classifier to predict buying intentions based on user behaviour within a large e-commerce website. In this work we compare traditional machine learning techniques with the most advanced deep learning approaches. We show that both Deep Belief Networks and Stacked Denoising auto-Encoders achieved a su…

Whole MILC learns brain disorder dynamics from unlabeled data.

Headless attacks bypass classification heads to fool transfer learning models.

Method transfers knowledge without label overlap, source data, or target architecture consistency.

Self-training outperforms pre-training on COCO object detection and segmentation datasets.

He et al. (2018) have called into question the utility of pre-training by showing that training from scratch can often yield similar performance to pre-training. We show that although pre-training may not improve performance on traditional classification metrics, it improves model robustness and uncertainty estimates. …

RoBERTa outperforms other pre-trained models in NER tasks.

We introduce the BriarPatch, a pixel-space intervention that obscures sensitive attributes from representations encoded in pre-trained classifiers. The patches encourage internal model representations not to encode sensitive information, which has the effect of pushing downstream predictors towards exhibiting demograph…

New framework explains how larger pre-trained models reduce downstream learning sample complexity.

We propose a Bayesian evidence framework to facilitate transfer learning from pre-trained deep convolutional neural networks (CNNs). Our framework is formulated on top of a least squares SVM (LS-SVM) classifier, which is simple and fast in both training and testing, and achieves competitive performance in practice. The…

VQShape learns interpretable time-series representations and achieves comparable performance to specialist models.

We study the robustness of image classifiers to temporal perturbations derived from videos. As part of this study, we construct two datasets, ImageNet-Vid-Robust and YTBB-Robust , containing a total 57,897 images grouped into 3,139 sets of perceptually similar images. Our datasets were derived from ImageNet-Vid and You…

Pre-training on different modalities improves Transformer performance in offline reinforcement learning.

The problem of detecting whether a test sample is from in-distribution (i.e., training distribution by a classifier) or out-of-distribution sufficiently different from it arises in many real-world machine learning applications. However, the state-of-art deep neural networks are known to be highly overconfident in their…

The recent surge of text-based online counseling applications enables us to collect and analyze interactions between counselors and clients. A dataset of those interactions can be used to learn to automatically classify the client utterances into categories that help counselors in diagnosing client status and predictin…

The study finds a theoretical bound for pre-training iterations needed for pruning to yield good subnetwork performance.

Unsupervised pre-training improves model generalization, but lacks theoretical understanding.

Adversarially robust models transfer better than standard models in image classification.

Differentially private ensemble classifiers adapt to data streams while protecting privacy.