Some smart companies are already making billions of dollars from their investments in AI, but these companies are taking the long-term view to achieve that. And the longer view companies take when working on AI, the more likely AI will exceed their expectations. These are some of the conclusions that emerged from leading executives at VB Summit last week, which focused on how AI is accelerating business results. Leaders from Microsoft, Facebook, IBM, and Intel joined a "Titans in AI" panel at the end of the Summit, and elaborated on some of these findings. Another key finding from the event is that personnel decisions are critical.

The digital telecommunications receiver is an important context for inference methodology, the key objective being to minimize the expected loss function in recovering the transmitted information. For that criterion, the optimal decision is the Bayesian minimum-risk estimator. However, the computational load of the Bayesian estimator is often prohibitive and, hence, efficient computational schemes are required. The design of novel schemes, striking new balances between accuracy and computational load, is the primary concern of this thesis. Two popular techniques, one exact and one approximate, will be studied. The exact scheme is a recursive one, namely the generalized distributive law (GDL), whose purpose is to distribute all operators across the conditionally independent (CI) factors of the joint model, so as to reduce the total number of operators required. In a novel theorem derived in this thesis, GDL, if applicable, will be shown to guarantee such a reduction in all cases. An associated lemma also quantifies this reduction. For practical use, two novel algorithms, namely the no-longer-needed (NLN) algorithm and the generalized form of the Markovian Forward-Backward (FB) algorithm, recursively factorizes and computes the CI factors of an arbitrary model, respectively. The approximate scheme is an iterative one, namely the Variational Bayes (VB) approximation, whose purpose is to find the independent (i.e. zero-order Markov) model closest to the true joint model in the minimum Kullback-Leibler divergence (KLD) sense. Despite being computationally efficient, this naive mean field approximation confers only modest performance for highly correlated models. A novel approximation, namely Transformed Variational Bayes (TVB), will be designed in the thesis in order to relax the zero-order constraint in the VB approximation, further reducing the KLD of the optimal approximation.

Heteroscedastic regression which considers varying noises across input domain has many applications in fields like machine learning and statistics. Here we focus on the heteroscedastic Gaussian process (HGP) regression which integrates the latent function and the noise together in a unified non-parametric Bayesian framework. Though showing flexible and powerful performance, HGP suffers from the cubic time complexity, which strictly limits its application to big data. To improve the scalability of HGP, we first develop a variational sparse inference algorithm, named VSHGP, to handle large-scale datasets. Furthermore, to enhance the model capability of capturing quick-varying features, we follow the Bayesian committee machine (BCM) formalism to distribute the learning over multiple local VSHGP experts with many inducing points, and aggregate their predictive distributions. The proposed distributed VSHGP (DVSHGP) (i) enables large-scale HGP regression via distributed computations, and (ii) achieves high model capability via localized experts and many inducing points. Superiority of the proposed DVSHGP as compared to existing large-scale heteroscedastic/homoscedastic GPs is then verified using a synthetic dataset and three real-world datasets.

Though insect-sized flying robots have been around for a while, none had been able to take untethered fight until now. Engineers at the University of Washington have revealed the RoboFly had taken its first untethered flaps, earlier this year, marking the first time a wireless flying robotic insect has flown. Now the man behind the project has revealed he hopes to have fully autonomous swarms roaming the skies within five years. RoboFly is only slightly heavier than a toothpick and is powered by an onboard circuit that converts the laser energy into enough electricity to operate its wings. Previously, the electronics the insects carried to power and control their wings were too heavy for the robots to fly with, meaning they had to remain connected to a wire.

DALLAS--(BUSINESS WIRE)--To solve humanity's most complex and demanding problems ranging from creating sustainable global food production and preventing infectious disease epidemics to ensuring the safety of our planet and natural resources is the HPC community's next grand challenge. HPC and AI are revolutionizing how we untangle and solve global threats and humanitarian crises. The SC18 plenary session will examine the potential for advanced computing to help mitigate human suffering and elevate our capacity to protect the most vulnerable. This plenary session will hear from innovators who are redefining how we predict and prevent humanitarian crises by leveraging advanced computing. The session is the kick-off event, which immediately precedes the Exhibitor Opening Gala.

Ultrasound elastography is gaining traction as an accessible and useful diagnostic tool for such things as cancer detection and differentiation as well as liver and thyroid disease diagnostics. Unfortunately, state of the art acoustic radiation force techniques, essential to promote this goal, are limited to high end ultrasound hardware due to high power requirements; are extremely sensitive to patient and sonographer motion; and generally suffer from low frame rates. Researchers have shown that pressure wave velocity possesses similar diagnostic abilities to shear wave velocity. Using pressure waves removes the need for generating shear waves, which in turn enables elasticity based diagnostic techniques on portable and low cost devices. However, current travel time tomography and full waveform inversion techniques for recovering pressure wave velocities require a full circumferential field of view. Focus based techniques, on the other hand, provide only localized measurements, are sensitive to the intermediate medium and require capturing multiple frames. In this paper, we present a single sided sound speed inversion solution using a fully convolutional deep neural network. We show that it is possible to invert for longitudinal sound speed in soft tissue at real time frame rates. For the computation, analysis is performed on channel data information from three diagonal plane waves. This is the first step towards a full waveform solver using a Deep Learning framework for the elastic and viscoelastic inverse problem.

Sequential learning of tasks using gradient descent leads to an unremitting decline in the accuracy of tasks for which training data is no longer available, termed catastrophic forgetting. Generative models have been explored as a means to approximate the distribution of old tasks and bypass storage of real data. Here we propose a cumulative closed-loop generator and embedded classifier using an AC-GAN architecture provided with external regularization by a small buffer. We evaluate incremental learning using a notoriously hard paradigm, single headed learning, in which each task is a disjoint subset of classes in the overall dataset, and performance is evaluated on all previous classes. First, we show that the variability contained in a small percentage of a dataset (memory buffer) accounts for a significant portion of the reported accuracy, both in multi-task and continual learning settings. Second, we show that using a generator to continuously output new images while training provides an up-sampling of the buffer, which prevents catastrophic forgetting and yields superior performance when compared to a fixed buffer. We achieve an average accuracy for all classes of 92.26% in MNIST and 76.15% in FASHION-MNIST after 5 tasks using GAN sampling with a buffer of only 0.17% of the entire dataset size. We compare to a network with regularization (EWC) which shows a deteriorated average performance of 29.19% (MNIST) and 26.5% (FASHION). The baseline of no regularization (plain gradient descent) performs at 99.84% (MNIST) and 99.79% (FASHION) for the last task, but below 3% for all previous tasks. Our method has very low long-term memory cost, the buffer, as well as negligible intermediate memory storage.

Smart data and the energy transition What is smart data and how can we use it to help us through the energy transition? In this episode, we talk to Dr Kirk Borne, Principal Data Scientist and Executive Advisor at Booz Allen Hamilton, about the advancements in technology that are aiding the transformation of the energy sector. In this latest episode, Kirk shares his views on how data is helping to drive energy efficiency and reduce climate change. Kirk explains why it is important to collect data from many different sensors to help us to fully understand the position on climate change; the importance of integration of data sources to break down the silos of information available; and how taking a prescriptive modelling approach will help us to achieve a better outcome for the planet. Finally, we talk to Kirk about AI, in particular Assisted Intelligence, and how humans and machines are working together to filter and understand the significant amount of data available.

A newly introduced method called Taylor-based Optimized Recursive Extended Exponential Smoothed Neural Networks Forecasting method is applied and extended in this study to forecast numerical values. Unlike traditional forecasting techniques which forecast only future values, our proposed method provides a new extension to correct the predicted values which is done by forecasting the estimated error. Experimental results demonstrated that the proposed method has a high accuracy both in training and testing data and outperform the state-of-the-art RNN models on Mackey-Glass, NARMA, Lorenz and Henon map datasets.

With Google announcing a new £19m fund for businesses using Artificial Intelligence (AI) to address environmental and social issues, edie looks at nine projects that have already delivered notable benefits for business. According to PwC, the global economy could see a potential contribution of $15.7trn from AI by 2030 Google has today (30 October) issued a rallying cry to the tech-savvy businesses of the world. The technology giant has launched a $25m (£19m) Impact Challenge, where organisations of all sizes can submit concepts for how AI can be used to alleviate and address key societal and environmental challenges. "We're looking for projects across a range of social impact domains and levels of technical expertise, from organisations that are experienced in AI to those with an idea for how they could be putting their data to better use," a Google blog post reads. Successful applicants will need to consider feasibility, scalability and responsibility of the AI concept, and will gain access to the funding pool and credit and consulting from Google Cloud.