Regularization in the optimization of deep neural networks is often critical to avoid undesirable over-fitting leading to better generalization of model. One of the most popular regularization algorithms is to impose L-2 penalty on the model parameters resulting in the decay of parameters, called weight-decay, and the decay rate is generally constant to all the model parameters in the course of optimization. In contrast to the previous approach based on the constant rate of weight-decay, we propose to consider the residual that measures dissimilarity between the current state of model and observations in the determination of the weight-decay for each parameter in an adaptive way, called adaptive weight-decay (AdaDecay) where the gradient norms are normalized within each layer and the degree of regularization for each parameter is determined in proportional to the magnitude of its gradient using the sigmoid function. We empirically demonstrate the effectiveness of AdaDecay in comparison to the state-of-the-art optimization algorithms using popular benchmark datasets: MNIST, Fashion-MNIST, and CIFAR-10 with conventional neural network models ranging from shallow to deep. The quantitative evaluation of our proposed algorithm indicates that AdaDecay improves generalization leading to better accuracy across all the datasets and models.

Facebook today announced the open source release of Deep Learning Recommendation Model (DLRM), a state-of-the-art AI model for serving up personalized results in production environments. DLRM can be found on GitHub, and implementations of the model are available for Facebook's PyTorch, Facebook's distributed learning framework Caffe2, and Glow C . Recommendation engines decide a lot of what people see today, whether it's content on social media sites like Facebook, ecommerce sites like Amazon, or even the first options you see on an Xbox. Last month, Amazon made its AI for the shopping recommendations system Personalize available on AWS. A paper by more than 20 Facebook AI researchers published on arXiv in late May explains how the model uses embedding tables that map categorical data to representations.

AI has got an unbelievable momentum in the past couple of years. The current intelligent frameworks have the capability of managing a lot of data and simplifying complicated calculations very fast. In any case, these are not the sentient machines. AI developers are trying to build up this feature in the future. And in the coming years, the AI framework will reach and surpass the performance of humans in solving different tasks.

Twitter @tarantulae 4. Uncertainty in Deep Learning - Christian S. Perone (2019) Uncertainties Bayesian Inference Deep Learning Variational Inference Ensembles Q&A Section I Uncertainties 5. Uncertainty in Deep Learning - Christian S. Perone (2019) Uncertainties Bayesian Inference Deep Learning Variational Inference Ensembles Q&A Knowing what you don't know It is correct, somebody might say, that (...) Socrates did not know anything; and it was indeed wisdom that they recognized their own lack of knowledge, (...).

This blog post is an overview of quantum machine learning written by the author of the paper Bayesian deep learning on a quantum computer. In it, we explore the application of machine learning in the quantum computing space. The authors of this paper hope that the results of the experiment help influence the future development of quantum machine learning. With no shortage of research problems, education programs, and demand for talent, machine learning is one of the hottest topics in technology today. Parallel to the success of learning algorithms, the development of quantum computing hardware has accelerated over the last few years.

This week's furor over FaceApp has largely centered on concerns that its Russian developers might be compelled to share the app's data with the Russian government, much as the Snowden disclosures illustrated the myriad ways in which American companies were compelled to disclose their private user data to the US government. Yet the reality is that this represents a mistaken understanding of just how the modern data trade works today and the simple fact that American universities and companies routinely make their data available to companies all across the world, including in Russia and China. In today's globalized world, data is just as globalized, with national borders no longer restricting the flow of our personal information - trend made worse by the data-hungry world of deep learning. Data brokers have long bought and sold our personal data in a shadowy world of international trade involving our most intimate and private information. The digital era has upended this explicit trade through the interlocking world of passive exchange through analytics services.

We present an approach to learn the dynamics of multiple objects from image sequences in an unsupervised way. We introduce a probabilistic model that first generate noisy positions for each object through a separate linear state-space model, and then renders the positions of all objects in the same image through a highly non-linear process. Such a linear representation of the dynamics enables us to propose an inference method that uses exact and efficient inference tools and that can be deployed to query the model in different ways without retraining.

We introduce Arena, a toolkit for multi-agent reinforcement learning (MARL) research. In MARL, it usually requires customizing observations, rewards and actions for each agent, changing cooperative-competitive agent-interaction, and playing with/against a third-party agent, etc. We provide a novel modular design, called Interface, for manipulating such routines in essentially two ways: 1) Different interfaces can be concatenated and combined, which extends the OpenAI Gym Wrappers concept to MARL scenarios. 2) During MARL training or testing, interfaces can be embedded in either wrapped OpenAI Gym compatible Environments or raw environment compatible Agents. We offer off-the-shelf interfaces for several popular MARL platforms, including StarCraft II, Pommerman, ViZDoom, Soccer, etc. The interfaces effectively support self-play RL and cooperative-competitive hybrid MARL. Also, Arena can be conveniently extended to your own favorite MARL platform.

Many real-world data sets, especially in biology, are produced by complex nonlinear dynamical systems. In this paper, we focus on brain calcium imaging (CaI) of different organisms (zebrafish and rat), aiming to build a model of joint activation dynamics in large neuronal populations, including the whole brain of zebrafish. We propose a new approach for capturing dynamics of temporal SVD components that uses the coupled (multivariate) van der Pol (VDP) oscillator, a nonlinear ordinary differential equation (ODE) model describing neural activity, with a new parameter estimation technique that combines variable projection optimization and stochastic search. We show that the approach successfully handles nonlinearities and hidden state variables in the coupled VDP. The approach is accurate, achieving 0.82 to 0.94 correlation between the actual and model-generated components, and interpretable, as VDP's coupling matrix reveals anatomically meaningful positive (excitatory) and negative (inhibitory) interactions across different brain subsystems corresponding to spatial SVD components. Moreover, VDP is comparable to (or sometimes better than) recurrent neural networks (LSTM) for (short-term) prediction of future brain activity; VDP needs less parameters to train, which was a plus on our small training data. Finally, the overall best predictive method, greatly outperforming both VDP and LSTM in short- and long-term predictive settings on both datasets, was the new hybrid VDP-LSTM approach that used VDP to simulate large domain-specific dataset for LSTM pretraining; note that simple LSTM data-augmentation via noisy versions of training data was much less effective.

Traffic signals serve to regulate the worst bottlenecks in highly populated areas but are not always very effective. Researchers at Penn State are hoping to use deep reinforcement learning to improve traffic signal efficiency in urban areas, thanks to a one-year, $22,443 Penn State Institute for CyberScience Seed Grant. Urban traffic congestion currently costs the U.S. economy $160 billion in lost productivity and causes 3.1 billion gallons of wasted fuel and 56 billion pounds of harmful CO2 emissions, according to the 2015 Urban Mobility Scorecard. Vikash Gayah, associate professor of civil engineering, and Zhenhui "Jessie" Li, associate professor of information sciences and technology, aim to tackle this issue by first identifying machine learning algorithms that will provide results consistent with traditional (theoretical) solutions for simple scenerios, and then building upon those algorithms by introducing complexities that cannot be readily addressed through traditional means. "Typically, we would go out and do traffic counts for an hour at certain peak times of day and that would determine signal timings for the next year, but not every day looks like that hour, and so we get inefficiency," Gayah said.