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
OptEmbed: Learning Optimal Embedding Table for Click-through Rate Prediction
Click-through rate (CTR) prediction model usually consists of three components: embedding table, feature interaction layer, and classifier. … (voir plus)Learning embedding table plays a fundamental role in CTR prediction from the view of the model performance and memory usage. The embedding table is a two-dimensional tensor, with its axes indicating the number of feature values and the embedding dimension, respectively. To learn an efficient and effective embedding table, recent works either assign various embedding dimensions for feature fields and reduce the number of embeddings respectively or mask the embedding table parameters. However, all these existing works cannot get an optimal embedding table. On the one hand, various embedding dimensions still require a large amount of memory due to the vast number of features in the dataset. On the other hand, decreasing the number of embeddings usually suffers from performance degradation, which is intolerable in CTR prediction. Finally, pruning embedding parameters will lead to a sparse embedding table, which is hard to be deployed. To this end, we propose an optimal embedding table learning framework OptEmbed, which provides a practical and general method to find an optimal embedding table for various base CTR models. Specifically, we propose pruning the redundant embeddings regarding corresponding features' importance by learnable pruning thresholds. Furthermore, we consider assigning various embedding dimensions as one single candidate architecture. To efficiently search the optimal embedding dimensions, we design a uniform embedding dimension sampling scheme to equally train all candidate architectures, meaning architecture-related parameters and learnable thresholds are trained simultaneously in one supernet. We then propose an evolution search method based on the supernet to find the optimal embedding dimensions for each field. Experiments on public datasets show that OptEmbed can learn a compact embedding table which can further improve the model performance.
2022-10-17
Proceedings of the 31st ACM International Conference on Information & Knowledge Management (publié)
Recent work on Graph Neural Networks has demonstrated that self-supervised pretraining can further enhance performance on downstream graph, … (voir plus)link, and node classification tasks. However, the efficacy of pretraining tasks has not been fully investigated for downstream large knowledge graph completion tasks. Using a contextualized knowledge graph embedding approach, we investigate five different pretraining signals, constructed using several graph algorithms and no external data, as well as their combination. We leverage the versatility of our Transformer-based model to explore graph structure generation pretraining tasks (i.e. path and k-hop neighborhood generation), typically inapplicable to most graph embedding methods. We further propose a new path-finding algorithm guided by information gain and find that it is the best-performing pretraining task across three downstream knowledge graph completion datasets. While using our new path-finding algorithm as a pretraining signal provides 2-3% MRR improvements, we show that pretraining on all signals together gives the best knowledge graph completion results. In a multitask setting that combines all pretraining tasks, our method surpasses the latest and strong performing knowledge graph embedding methods on all metrics for FB15K-237, on MRR and Hit@1 for WN18RRand on MRR and hit@10 for JF17K (a knowledge hypergraph dataset).
Adaptive methods are a crucial component widely used for training generative adversarial networks (GANs). While there has been some work to … (voir plus)pinpoint the “marginal value of adaptive methods” in standard tasks, it remains unclear why they are still critical for GAN training. In this paper, we formally study how adaptive methods help train GANs; inspired by the grafting method proposed in Agarwal et al. (2020), we separate the magnitude and direction components of the Adam updates, and graft them to the direction and magnitude of SGDA updates respectively. By considering an update rule with the magnitude of the Adam update and the normalized direction of SGD, we empirically show that the adaptive magnitude of Adam is key for GAN training. This motivates us to have a closer look at the class of normalized stochastic gradient descent ascent (nSGDA) methods in the context of GAN training. We propose a synthetic theoretical framework to compare the performance of nSGDA and SGDA for GAN training with neural networks. We prove that in that setting, GANs trained with nSGDA recover all the modes of the true distribution, whereas the same networks trained with SGDA (and any learning rate configuration) suffer from mode collapse. The critical insight in our analysis is that normalizing the gradients forces the discriminator and generator to be updated at the same pace. We also experimentally show that for several datasets, Adam’s performance can be recovered with nSGDA methods.
Strong static type systems help programmers eliminate many errors without much burden of supplying type annotations. However, this flexibili… (voir plus)ty makes it highly non-trivial to diagnose ill-typed programs, especially for novice programmers. Compared to classic constraint solving and optimization-based approaches, the data-driven approach has shown great promise in identifying the root causes of type errors with higher accuracy. Instead of relying on hand-engineered features, this work explores natural language models for type error localization, which can be trained in an end-to-end fashion without requiring any features. We demonstrate that, for novice type error diagnosis, the language model-based approach significantly outperforms the previous state-of-the-art data-driven approach. Specifically, our model could predict type errors correctly 62% of the time, outperforming the state-of-the-art Nate's data-driven model by 11%, in a more rigorous accuracy metric. Furthermore, we also apply structural probes to explain the performance difference between different language models.
Our understanding of the changes in functional brain organization in autism is hampered by the extensive heterogeneity that characterizes th… (voir plus)is neurodevelopmental disorder. Data driven clustering offers a straightforward way to decompose autism heterogeneity into subtypes of connectivity and promises an unbiased framework to investigate behavioral symptoms and causative genetic factors. Yet, the robustness and generalizability of functional connectivity subtypes is unknown. Here, we show that a simple hierarchical cluster analysis can robustly relate a given individual and brain network to a connectivity subtype, but that continuous assignments are more robust than discrete ones. We also found that functional connectivity subtypes are moderately associated with the clinical diagnosis of autism, and these associations generalize to independent replication data. We explored systematically 18 different brain networks as we expected them to associate with different behavioral profiles as well as different key regions. Contrary to this prediction, autism functional connectivity subtypes converged on a common topography across different networks, consistent with a compression of the primary gradient of functional brain organization, as previously reported in the literature. Our results support the use of data driven clustering as a reliable data dimensionality reduction technique, where any given dimension only associates moderately with clinical manifestations.
Our understanding of the changes in functional brain organization in autism is hampered by the extensive heterogeneity that characterizes th… (voir plus)is neurodevelopmental disorder. Data driven clustering offers a straightforward way to decompose autism heterogeneity into subtypes of connectivity and promises an unbiased framework to investigate behavioral symptoms and causative genetic factors. Yet, the robustness and generalizability of functional connectivity subtypes is unknown. Here, we show that a simple hierarchical cluster analysis can robustly relate a given individual and brain network to a connectivity subtype, but that continuous assignments are more robust than discrete ones. We also found that functional connectivity subtypes are moderately associated with the clinical diagnosis of autism, and these associations generalize to independent replication data. We explored systematically 18 different brain networks as we expected them to associate with different behavioral profiles as well as different key regions. Contrary to this prediction, autism functional connectivity subtypes converged on a common topography across different networks, consistent with a compression of the primary gradient of functional brain organization, as previously reported in the literature. Our results support the use of data driven clustering as a reliable data dimensionality reduction technique, where any given dimension only associates moderately with clinical manifestations.
Protective effectiveness of prior SARS-CoV-2 infection and hybrid immunity against Omicron infection and severe disease: a systematic review and meta-regression
Background We aimed to systematically review the magnitude and duration of the protective effectiveness of prior infection (PE) and hybrid i… (voir plus)mmunity (HE) against Omicron infection and severe disease. Methods We searched pre-print and peer-reviewed electronic databases for controlled studies from January 1, 2020, to June 1, 2022. Risk of bias (RoB) was assessed using the Risk of Bias In Non-Randomized Studies of Interventions (ROBINS-I)-Tool. We used random-effects meta-regression to estimate the magnitude of protection at 1-month intervals and the average change in protection since the last vaccine dose or infection from 3 months to 6 or 12 months. We compared our estimates of PE and HE to previously published estimates of the magnitude and durability of vaccine effectiveness (VE) against Omicron. Findings Eleven studies of prior infection and 15 studies of hybrid immunity were included. For prior infection, there were 97 estimates (27 at moderate RoB and 70 at serious RoB), with the longest follow up at 15 months. PE against hospitalization or severe disease was 82.5% [71.8-89.7%] at 3 months, and 74.6% [63.1-83.5%] at 12 months. PE against reinfection was 65.2% [52.9-75.9%] at 3 months, and 24.7% [16.4-35.5%] at 12 months. For HE, there were 153 estimates (78 at moderate RoB and 75 at serious RoB), with the longest follow up at 11 months for primary series vaccination and 4 months for first booster vaccination. Against hospitalization or severe disease, HE involving either primary series vaccination or first booster vaccination was consistently >95% for the available follow up. Against reinfection, HE involving primary series vaccination was 69.0% [58.9-77.5%] at 3 months after the most recent infection or vaccination, and 41.8% [31.5-52.8%] at 12 months, while HE involving first booster vaccination was 68.6% [58.8-76.9%] at 3 months, and 46.5% [36.0-57.3%] at 6 months. Against hospitalization or severe disease at 6 months, hybrid immunity with first booster vaccination (effectiveness 95.3% [81.9-98.9%]) or with primary series alone (96.5% [90.2-98.8%]) provided significantly greater protection than prior infection alone (80.1% [70.3-87.2%]), first booster vaccination alone (76.7% [72.5-80.4%]), or primary series alone (64.6% [54.5-73.6%]). Results for protection against reinfection were similar. Interpretation Prior infection and hybrid immunity both provided greater and more sustained protection against Omicron than vaccination alone. All protection estimates waned quickly against infection but remained high for hospitalisation or severe disease. Individuals with hybrid immunity had the highest magnitude and durability of protection against all outcomes, reinforcing the global imperative for vaccination.
A Reproducible and Realistic Evaluation of Partial Domain Adaptation Methods
Unsupervised Domain Adaptation (UDA) aims at classifying unlabeled target images leveraging source labeled ones. In this work, we consider t… (voir plus)he Partial Domain Adaptation (PDA) variant, where we have extra source classes not present in the target domain. Most successful algorithms use model selection strategies that rely on target labels to find the best hyper-parameters and/or models along training. However, these strategies violate the main assumption in PDA: only unlabeled target domain samples are available. Moreover, there are also inconsistencies in the experimental settings - architecture, hyper-parameter tuning, number of runs - yielding unfair comparisons. The main goal of this work is to provide a realistic evaluation of PDA methods with the different model selection strategies under a consistent evaluation protocol. We evaluate 7 representative PDA algorithms on 2 different real-world datasets using 7 different model selection strategies. Our two main findings are: (i) without target labels for model selection, the accuracy of the methods decreases up to 30 percentage points; (ii) only one method and model selection pair performs well on both datasets. Experiments were performed with our PyTorch framework, BenchmarkPDA, which we open source.
