Le traitement du langage naturel à l'ère de l'IA générative
Rejoignez-nous à Mila en octobre pour un atelier de trois jour visant à explorer le potentiel de transformation des technologies langagières et leurs implications pour la société.
Ce programme est conçu pour fournir aux professionnel·le·s travaillant dans le domaine de la politique une compréhension fondamentale de la technologie de l'IA.
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
Modular Networks for Validating Community Detection Algorithms
How can we accurately compare different community detection algorithms? These algorithms cluster nodes in a given network, and their perform… (voir plus)ance is often validated on benchmark networks with explicit ground-truth communities. Given the lack of cluster labels in real-world networks, a model that generates realistic networks is required for accurate evaluation of these algorithm. In this paper, we present a simple, intuitive, and flexible benchmark generator to generate intrinsically modular networks for community validation. We show how the generated networks closely comply with the characteristics observed for real networks; whereas their characteristics could be directly controlled to match wide range of real world networks. We further show how common community detection algorithms rank differently when being evaluated on these benchmarks compared to current available alternatives.
Principal component analysis (PCA) is one of the most powerful tools in machine learning. The simplest method for PCA, the power iteration, … (voir plus)requires O ( 1 / Δ ) full-data passes to recover the principal component of a matrix with eigen-gap Δ. Lanczos, a significantly more complex method, achieves an accelerated rate of O ( 1 / Δ ) passes. Modern applications, however, motivate methods that only ingest a subset of available data, known as the stochastic setting. In the online stochastic setting, simple algorithms like Oja's iteration achieve the optimal sample complexity O ( σ 2 / Δ 2 ) . Unfortunately, they are fully sequential, and also require O ( σ 2 / Δ 2 ) iterations, far from the O ( 1 / Δ ) rate of Lanczos. We propose a simple variant of the power iteration with an added momentum term, that achieves both the optimal sample and iteration complexity. In the full-pass setting, standard analysis shows that momentum achieves the accelerated rate, O ( 1 / Δ ) . We demonstrate empirically that naively applying momentum to a stochastic method, does not result in acceleration. We perform a novel, tight variance analysis that reveals the "breaking-point variance" beyond which this acceleration does not occur. By combining this insight with modern variance reduction techniques, we construct stochastic PCA algorithms, for the online and offline setting, that achieve an accelerated iteration complexity O ( 1 / Δ ) . Due to the embarassingly parallel nature of our methods, this acceleration translates directly to wall-clock time if deployed in a parallel environment. Our approach is very general, and applies to many non-convex optimization problems that can now be accelerated using the same technique.
Automatic differentiation in ML: Where we are and where we should be going
Bart van Merriënboer
Olivier Breuleux
Arnaud Bergeron
Pascal Lamblin
We review the current state of automatic differentiation (AD) for array programming in machine learning (ML), including the different approa… (voir plus)ches such as operator overloading (OO) and source transformation (ST) used for AD, graph-based intermediate representations for programs, and source languages. Based on these insights, we introduce a new graph-based intermediate representation (IR) which specifically aims to efficiently support fully-general AD for array programming. Unlike existing dataflow programming representations in ML frameworks, our IR naturally supports function calls, higher-order functions and recursion, making ML models easier to implement. The ability to represent closures allows us to perform AD using ST without a tape, making the resulting derivative (adjoint) program amenable to ahead-of-time optimization using tools from functional language compilers, and enabling higher-order derivatives. Lastly, we introduce a proof of concept compiler toolchain called Myia which uses a subset of Python as a front end.
Minimizing a function over an intersection of convex sets is an important task in optimization that is often much more challenging than mini… (voir plus)mizing it over each individual constraint set. While traditional methods such as Frank-Wolfe (FW) or proximal gradient descent assume access to a linear or quadratic oracle on the intersection, splitting techniques take advantage of the structure of each sets, and only require access to the oracle on the individual constraints. In this work, we develop and analyze the Frank-Wolfe Augmented Lagrangian (FW-AL) algorithm, a method for minimizing a smooth function over convex compact sets related by a "linear consistency" constraint that only requires access to a linear minimization oracle over the individual constraints. It is based on the Augmented Lagrangian Method (ALM), also known as Method of Multipliers, but unlike most existing splitting methods, it only requires access to linear (instead of quadratic) minimization oracles. We use recent advances in the analysis of Frank-Wolfe and the alternating direction method of multipliers algorithms to prove a sublinear convergence rate for FW-AL over general convex compact sets and a linear convergence rate for polytopes.
Deep neural networks have been displaying superior performance over traditional supervised classifiers in text classification. They learn to… (voir plus) extract useful features automatically when sufficient amount of data is presented. However, along with the growth in the number of documents comes the increase in the number of categories, which often results in poor performance of the multiclass classifiers. In this work, we use external knowledge in the form of topic category taxonomies to aide the classification by introducing a deep hierarchical neural attention-based classifier. Our model performs better than or comparable to state-of-the-art hierarchical models at significantly lower computational cost while maintaining high interpretability.
2018-01-01
Conference on Empirical Methods in Natural Language Processing (publié)
Since the first MICCAI grand challenge was organized in 2007 [1], the impact of biomedical image analysis challenges on both the research fi… (voir plus)eld as well as on individual careers has been steadily growing. For example, the acceptance of a journal article today often depends on the performance of a new algorithm being assessed against the state-ofthe-art work on publicly available challenge datasets. Furthermore, the results are also important for the individuals scientific careers as well as the potential that algorithms can be translated into clinical practice. Yet, while the publication of papers in scientific journals and prestigious conferences, such as MICCAI, undergoes strict quality control, the design and organization of challenges do not. To investigate the effect of common practice, we have formed an international initiative dedicated to analyzing and improving a variety of aspects related to biomedical challenge design, execution and reporting [2]. In the first part of our abstract presentation at LABELS workshop, we are going to present some of the major pitfalls related to biomedical image analysis challenges today. Specifically, we will look at the following research questions: RQ1: How robust are challenge rankings? What is the effect of – the specific test cases used? – the specific metric variant(s) applied? – the rank aggregation method chosen (e.g. aggregation of metric values with the mean vs median)? ? Shared first/senior authors.
Graph Weighted Models (GWMs) have recently been proposed as a natural generalization of weighted automata over strings and trees to arbitrar… (voir plus)y families of labeled graphs (and hypergraphs). A GWM generically associates a labeled graph with a tensor network and computes a value by successive contractions directed by its edges. In this paper, we consider the problem of learning GWMs defined over the graph family of pictures (or 2-dimensional words). As a proof of concept, we consider regression and classification tasks over the simple Bars & Stripes and Shifting Bits picture languages and provide an experimental study investigating whether these languages can be learned in the form of a GWM from positive and negative examples using gradient-based methods. Our results suggest that this is indeed possible and that investigating the use of gradient-based methods to learn picture series and functions computed by GWMs over other families of graphs could be a fruitful direction.
In this paper, we develop algorithmic approaches for a recently defined class of games, the integer programming games. Two general methods t… (voir plus)o approximate an equilibrium are presented and enhanced in order to improve their practical efficiency. Their performance is analysed through computational experiments in a knapsack game and a competitive lot-sizing game. To the best of our knowledge, this is the first time that equilibria computation methods for general integer programming games are build and computationally tested.
The loss function of deep networks is known to be non-convex but the precise nature of this nonconvexity is still an active area of research… (voir plus). In this work, we study the loss landscape of deep networks through the eigendecompositions of their Hessian matrix. In particular, we examine how important the negative eigenvalues are and the benefits one can observe in handling them appropriately.