In many real world applications, data are characterized by a complex structure, that can be naturally encoded as a graph. In the last years, the popularity of deep learning techniques has renewed the interest in neural models able to process complex patterns. In particular, inspired by the Graph Neural Network (GNN) model, different architectures have been proposed to extend the original GNN scheme. GNNs exploit a set of state variables, each assigned to a graph node, and a diffusion mechanism of the states among neighbor nodes, to implement an iterative procedure to compute the fixed point of the (learnable) state transition function. In this paper, we propose a novel approach to the state computation and the learning algorithm for GNNs, based on a constraint optimisation task solved in the Lagrangian framework. The state convergence procedure is implicitly expressed by the constraint satisfaction mechanism and does not require a separate iterative phase for each epoch of the learning procedure. In fact, the computational structure is based on the search for saddle points of the Lagrangian in the adjoint space composed of weights, neural outputs (node states), and Lagrange multipliers. The proposed approach is compared experimentally with other popular models for processing graphs.

ABSTRACT We consider statistical learning problems, when the distribution P ′ of the training observations Z ′ 1,..., Z′ n differs from the distribution P involved in the risk one seeks to minimize (referred to as the test distribution) but is still defined on the same measurable space as P and dominates it. In the unrealistic case where the likelihood ratio Φ(z) dP/dP ′ (z) is known, one may straightforwardly extends the Empirical Risk Minimization (ERM) approach to this specific transfer learning setup using the same idea as that behind Importance Sampling, by minimizing a weighted version of the empirical risk functional computed from the'biased' training data Zi ′ with weights Φ(Zi ′). Although the importance function Φ(z) is generally unknown in practice, we show that, in various situations frequently encountered in practice, it takes a simple form and can be directly estimated from the Zi ′ 's and some auxiliary information on the statistical population P. By means of linearization techniques, we then prove that the generalization capacity of the approach aforementioned is preserved when plugging the resulting estimates of the Φ(Zi ′)'s into the weighted empirical risk. Beyond these theoretical guarantees, numerical results provide strong empirical evidence of the relevance of the approach promoted in this article. Keywords: Statistical Learning Theory, Importance Sampling, Transfer Learning. 1 Introduction Prediction problems are of major importance in statistical learning. The main paradigm of predictive learning is Empirical Risk Minimization (ERM in abbreviated form), see e.g. In the standard setup, Z is a random variable (r.v. in short) that takes its values in a feature space Z with distribution P, Θ is a parameter space and l: Θ Z R is a (measurable) loss function. The risk is then defined by: θ Θ, R P (θ) E P [l(θ, Z)], (1) and more generally for any measure Q on Z: R Q (θ) l(θ, z)dQ(z).

State-of-the-art, high capacity deep neural networks not only require large amounts of labelled training data, they are also highly susceptible to label errors in this data, typically resulting in large efforts and costs and therefore limiting the applicability of deep learning. To alleviate this issue, we propose a novel meta training and labelling scheme that is able to use inexpensive unlabelled data by taking advantage of the generalization power of deep neural networks. We show experimentally that by solely relying on one network architecture and our proposed scheme of iterative training and prediction steps, both label quality and resulting model accuracy can be improved significantly. Our method achieves state-of-the-art results, while being architecture agnostic and therefore broadly applicable. Compared to other methods dealing with erroneous labels, our approach does neither require another network to be trained, nor does it necessarily need an additional, highly accurate reference label set. Instead of removing samples from a labelled set, our technique uses additional sensor data without the need for manual labelling.

Real-time traffic accident forecasting is increasingly important for public safety and urban management (e.g., real-time safe route planning and emergency response deployment). Previous works on accident forecasting are often performed on hour levels, utilizing existed neural networks with static region-wise correlations taken into account. However, it is still challenging when the granularity of forecasting step improves as the highly dynamic nature of road network and inherent rareness of accident records in one training sample, which leads to biased results and zero-inflated issue. In this work, we propose a novel framework RiskOracle, to improve the prediction granularity to minute levels. Specifically, we first transform the zero-risk values in labels to fit the training network. Then, we propose the Differential Time-varying Graph neural network (DTGN) to capture the immediate changes of traffic status and dynamic inter-subregion correlations. Furthermore, we adopt multi-task and region selection schemes to highlight citywide most-likely accident subregions, bridging the gap between biased risk values and sporadic accident distribution. Extensive experiments on two real-world datasets demonstrate the effectiveness and scalability of our RiskOracle framework.

Recent efforts in Dialogue State Tracking (DST) for task-oriented dialogues have progressed toward open-vocabulary or generation-based approaches where the models can generate slot value candidates from the dialogue history itself. These approaches have shown good performance gain, especially in complicated dialogue domains with dynamic slot values. However, they fall short in two aspects: (1) they do not allow models to explicitly learn signals across domains and slots to detect potential dependencies among (domain, slot) pairs; and (2) existing models follow auto-regressive approaches which incur high time cost when the dialogue evolves over multiple domains and multiple turns. In this paper, we propose a novel framework of Non-Autoregressive Dialog State Tracking (NADST) which can factor in potential dependencies among domains and slots to optimize the models towards better prediction of dialogue states as a complete set rather than separate slots. In particular, the non-autoregressive nature of our method not only enables decoding in parallel to significantly reduce the latency of DST for real-time dialogue response generation, but also detect dependencies among slots at token level in addition to slot and domain level. Our empirical results show that our model achieves the state-of-the-art joint accuracy across all domains on the MultiWOZ 2.1 corpus, and the latency of our model is an order of magnitude lower than the previous state of the art as the dialogue history extends over time.

Recent research in artificial intelligence and machine learning has largely emphasized general-purpose learning and ever-larger training sets and more and more compute. In contrast, I propose a hybrid, knowledge-driven, reasoning-based approach, centered around cognitive models, that could provide the substrate for a richer, more robust AI than is currently possible.

Tesla and SpaceX CEO, Elon Musk, says that AI like the one his companies make should be better regulated. Musk's opinion on the dangers of letting AI proliferate unfettered was prompted by a report published in MIT Technology Review about changing company culture at OpenAI, a technology company that helps develop new AI. Elon Musk formerly helmed the company but left due to conflicts of interest. The report claims that OpenAI has shifted from its goal of equitably distributing AI technology to a more secretive, funding-driven company. 'OpenAI should be more open imo,' he tweeted.

Heart disease could mean range of different conditions that could affect your heart. It is one of the most complex disease to predict given number of factors in your body that can potentially lead to it. Identifying and predicting it poses a great deal of challenge for doctors and researchers alike. I will attempt to take a stab at this problem using machine learning with the public dataset thats made available here at UCI Machine Learning Repository. There are 303 records in the dataset and contains 14 continuous attributes.

The first GANs paper had just come out two years before we started working on the second edition, but we weren't sure of its relevance. However, GANs have evolved into one of the hottest and most widely used deep learning techniques. People use them for creating artwork, colorizing and improving the quality of photos, and to recreate old video game textures in higher resolutions. It goes without saying that an introduction to GANs was long overdue. Another important machine learning topic not included in previous editions is reinforcement learning, which has received a massive boost in attention recently. Thanks to impressive projects such as DeepMind's AlphaGo and AlphaGo Zero, reinforcement learning has received extensive news coverage.

Deep learning is on the rise in the recycling industry, according to the new e-book, Harnessing the Potential of AI, from Tomra Sorting Recycling, a sensor-based sorting technology company with headquarters in Norway. Dispelling a common misconception about AI, this latest e-book chronicles the 30-year history and contributions AI has already made to the recycling industry, as well the bright future that lies ahead, according to a Tomra news release. "In the months and years to come, those in the recycling industry will hear much more about deep learning, a powerful component of artificial intelligence," says Daniel Bender, technical manager, deep learning for Tomra Sorting. "Deep learning shows the promise of providing solutions for the industry's most complex sorting challenges. Recycling operations at the forefront of using AI to sort material stand to gain a significant advantage over companies who do not."