Nous utilisons des témoins pour analyser le trafic et l’utilisation de notre site web, afin de personnaliser votre expérience. Vous pouvez désactiver ces technologies à tout moment, mais cela peut restreindre certaines fonctionnalités du site. Consultez notre Politique de protection de la vie privée pour en savoir plus.
Paramètre des cookies
Vous pouvez activer et désactiver les types de cookies que vous souhaitez accepter. Cependant certains choix que vous ferez pourraient affecter les services proposés sur nos sites (ex : suggestions, annonces personnalisées, etc.).
Cookies essentiels
Ces cookies sont nécessaires au fonctionnement du site et ne peuvent être désactivés. (Toujours actif)
Cookies analyse
Acceptez-vous l'utilisation de cookies pour mesurer l'audience de nos sites ?
Multimedia Player
Acceptez-vous l'utilisation de cookies pour afficher et vous permettre de regarder les contenus vidéo hébergés par nos partenaires (YouTube, etc.) ?
Publications
A Unified Few-Shot Classification Benchmark to Compare Transfer and Meta Learning Approaches
Meta and transfer learning are two successful families of approaches to few-shot 1 learning. Despite highly related goals, state-of-the-art … (voir plus)advances in each family are 2 measured largely in isolation of each other. As a result of diverging evaluation 3 norms, a direct or thorough comparison of different approaches is challenging. 4 To bridge this gap, we introduce a few-shot classification evaluation protocol 5 named VTAB+MD with the explicit goal of facilitating sharing of insights from 6 each community. We demonstrate its accessibility in practice by performing a 7 cross-family study of the best transfer and meta learners which report on both a 8 large-scale meta-learning benchmark (Meta-Dataset, MD), and a transfer learning 9 benchmark (Visual Task Adaptation Benchmark, VTAB). We find that, on average, 10 large-scale transfer methods (Big Transfer, BiT) outperform competing approaches 11 on MD, even when trained only on ImageNet. In contrast, meta-learning approaches 12 struggle to compete on VTAB when trained and validated on MD. However, BiT 13 is not without limitations, and pushing for scale does not improve performance 14 on highly out-of-distribution MD tasks. We hope that this work contributes to 15 accelerating progress on few-shot learning research. 16
Few-shot classification aims to recognize unseen classes when presented with only a small number of samples. We consider the problem of mult… (voir plus)i-domain few-shot image classification, where unseen classes and examples come from diverse data sources. This problem has seen growing interest and has inspired the development of benchmarks such as Meta-Dataset. A key challenge in this multi-domain setting is to effectively integrate the feature representations from the diverse set of training domains. Here, we propose a Universal Representation Transformer (URT) layer, that meta-learns to leverage universal features for few-shot classification by dynamically re-weighting and composing the most appropriate domain-specific representations. In experiments, we show that URT sets a new state-of-the-art result on Meta-Dataset. Specifically, it achieves top-performance on the highest number of data sources compared to competing methods. We analyze variants of URT and present a visualization of the attention score heatmaps that sheds light on how the model performs cross-domain generalization.
Discrete-time diffusion-based generative models and score matching methods have shown promising results in modeling high-dimensional image d… (voir plus)ata. Recently, Song et al. (2021) show that diffusion processes that transform data into noise can be reversed via learning the score function, i.e. the gradient of the log-density of the perturbed data. They propose to plug the learned score function into an inverse formula to define a generative diffusion process. Despite the empirical success, a theoretical underpinning of this procedure is still lacking. In this work, we approach the (continuous-time) generative diffusion directly and derive a variational framework for likelihood estimation, which includes continuous-time normalizing flows as a special case, and can be seen as an infinitely deep variational autoencoder. Under this framework, we show that minimizing the score-matching loss is equivalent to maximizing a lower bound of the likelihood of the plug-in reverse SDE proposed by Song et al. (2021), bridging the theoretical gap.
For a natural language understanding bench-001 mark to be useful in research, it has to con-002 sist of examples that are diverse and diffi… (voir plus)-003 cult enough to discriminate among current and 004 near-future state-of-the-art systems. However, 005 we do not yet know how best to select pas-006 sages to collect a variety of challenging exam-007 ples. In this study, we crowdsource multiple-008 choice reading comprehension questions for 009 passages taken from seven qualitatively dis-010 tinct sources, analyzing what attributes of pas-011 sages contribute to the difficulty and question 012 types of the collected examples. To our sur-013 prise, we find that passage source, length, and 014 readability measures do not significantly affect 015 question difficulty. Through our manual anno-016 tation of seven reasoning types, we observe 017 several trends between passage sources and 018 reasoning types, e.g., logical reasoning is more 019 often required in questions written for techni-020 cal passages. These results suggest that when 021 creating a new benchmark dataset, selecting a 022 diverse set of passages can help ensure a di-023 verse range of question types, but that passage 024 difficulty need not be a priority. 025
Where Did You Learn That From? Surprising Effectiveness of Membership Inference Attacks Against Temporally Correlated Data in Deep Reinforcement Learning
Neurons in the dorsal visual pathway of the mammalian brain are selective for motion stimuli, with the complexity of stimulus representation… (voir plus)s increasing along the hierarchy. This progression is similar to that of the ventral visual pathway, which is well characterized by artificial neural networks (ANNs) optimized for object recognition. In contrast, there are no image-computable models of the dorsal stream with comparable explanatory power. We hypothesized that the properties of dorsal stream neurons could be explained by a simple learning objective: the need for an organism to orient itself during self-motion. To test this hypothesis, we trained a 3D ResNet to predict an agent’s self-motion parameters from visual stimuli in a simulated environment. We found that the responses in this network accounted well for the selectivity of neurons in a large database of single-neuron recordings from the dorsal visual stream of non-human primates. In contrast, ANNs trained on an action recognition dataset through supervised or self-supervised learning could not explain responses in the dorsal stream, despite also being trained on naturalistic videos with moving objects. These results demonstrate that an ecologically relevant cost function can account for dorsal stream properties in the primate brain.
The human pineal gland regulates day‐night dynamics of multiple physiological processes, especially through the secretion of melatonin. Us… (voir plus)ing mass‐spectrometry‐based proteomics and dedicated analysis tools, we identify proteins in the human pineal gland and analyze systematically their variation throughout the day and compare these changes in the pineal proteome between control specimens and donors diagnosed with autism. Results reveal diverse regulated clusters of proteins with, among others, catabolic carbohydrate process and cytoplasmic membrane‐bounded vesicle‐related proteins differing between day and night and/or control versus autism pineal glands. These data show novel and unexpected processes happening in the human pineal gland during the day/night rhythm as well as specific differences between autism donor pineal glands and those from controls.
In this work, we propose Bijective-Contrastive Estimation (BCE), a classification-based learning criterion for energy-based models. We gener… (voir plus)ate a collection of contrasting distributions using bijections, and solve all the classification problems between the original data distribution and the distributions induced by the bijections using a classifier parameterized by an energy model. We show that if the classification objective is minimized, the energy function will uniquely recover the data density up to a normalizing constant. This has the benefit of not having to explicitly specify a contrasting distribution, like noise contrastive estimation. Experimentally, we demonstrate that the proposed method works well on 2D synthetic datasets. We discuss the difficulty in high dimensional cases, and propose potential directions to explore for future work.
Autism spectrum disorder (ASD) is commonly understood as a network disorder, yet case-control analyses against typically-developing controls… (voir plus) (TD) have yielded somewhat inconsistent patterns of results. The current work was centered on a novel approach to profile functional network idiosyncrasy, the inter-individual variability in the association between functional network organization and brain anatomy, and we tested the hypothesis that idiosyncrasy contributes to connectivity alterations in ASD. Studying functional network idiosyncrasy in a multi-centric dataset with 157 ASD and 172 TD, our approach revealed higher idiosyncrasy in ASD in the default mode, somatomotor and attention networks together with reduced idiosyncrasy in the lateral temporal lobe. Idiosyncrasy was found to increase with age in both ASD and TD, and was significantly correlated with symptom severity in the former group. Association analysis with structural and molecular brain features indicated that patterns of functional network idiosyncrasy were not correlated with ASD-related cortical thickness alterations, but closely with the spatial expression patterns of intracortical ASD risk genes. In line with our main hypothesis, we could demonstrate that idiosyncrasy indeed plays a strong role in the manifestation of connectivity alterations that are measurable with conventional case-control designs and may, thus, be a principal driver of inconsistency in the autism connectomics literature. These findings support important interactions between the heterogeneity of individuals with an autism diagnosis and group-level functional signatures, and help to consolidate prior research findings on the highly variable nature of the functional connectome in ASD. Our study promotes idiosyncrasy as a potential individualized diagnostic marker of atypical brain network development.
The computational prediction of transcription factor binding sites remains a challenging problems in bioinformatics, despite significant met… (voir plus)hodological developments from the field of machine learning. Such computational models are essential to help interpret the non-coding portion of human genomes, and to learn more about the regulatory mechanisms controlling gene expression. In parallel, massive genome sequencing efforts have produced assembled genomes for hundred of vertebrate species, but this data is underused. We present PhyloReg, a new semi-supervised learning approach that can be used for a wide variety of sequence-to-function prediction problems, and that takes advantage of hundreds of millions of years of evolution to regularize predictors and improve accuracy. We demonstrate that PhyloReg can be used to better train a previously proposed deep learning model of transcription factor binding. Simulation studies further help delineate the benefits of the a pproach. G ains in prediction accuracy are obtained over a broad set of transcription factors and cell types.
2020-12-16
2020 IEEE International Conference on Bioinformatics and Biomedicine (BIBM) (publié)
BACKGROUND
True evidence-informed decision making in public health relies on incorporating evidence from a number of sources in addition to… (voir plus) traditional scientific evidence. Lack of access to these types of data, as well as ease of use and interpretability of scientific evidence contribute to limited uptake of evidence-informed decision making in practice. An electronic evidence system that includes multiple sources of evidence and potentially novel computational processing approaches or artificial intelligence holds promise as a solution to overcoming barriers to evidence-informed decision making in public health.
OBJECTIVE
To understand the needs and preferences for an electronic evidence system among public health professionals in Canada.
METHODS
An invitation to participate in an anonymous online survey was distributed via listservs of two Canadian public health organizations. Eligible participants were English or French speaking individuals currently working in public health. The survey contained both multiple choice and open-ended questions about needs and preferences relevant to an electronic evidence system. Quantitative responses were analyzed to explore differences by public health role. Inductive and deductive analysis methods were used to code and interpret the qualitative data. Ethics review was not required by the host institution.
RESULTS
Respondents (n = 371) were heterogeneous, spanning organizations, positions, and areas of practice within public health. Nearly all (98.0%) respondents indicated that an electronic evidence system would support their work. Respondents had high preferences for local contextual data, research and intervention evidence, and information about human and financial resources. Qualitative analyses identified a number of concerns, needs, and suggestions for development of such a system. Concerns ranged from personal use of such a system, to the ability of their organization to use such a system. Identified needs spanned the different sources of evidence including local context, research and intervention evidence, and resources and tools. Additional suggestions were identified to improve system usability.
CONCLUSIONS
Canadian public health professionals have positive perceptions towards an electronic evidence system that would bring together evidence from the local context, scientific research, and resources. Elements were also identified to increase the usability of an electronic evidence system.
