The reparameterization trick is widely used in variational inference as it yields more accurate estimates of the gradient of the variational objective than alternative approaches such as the score function method. Although there is overwhelming empirical evidence in the literature showing its success, there is relatively little research exploring why the reparameterization trick is so effective. We explore this under the idealized assumptions that the variational approximation is a mean-field Gaussian density and that the log of the joint density of the model parameters and the data is a quadratic function that depends on the variational mean. From this, we show that the marginal variances of the reparameterization gradient estimator are smaller than those of the score function gradient estimator. We apply the result of our idealized analysis to real-world examples.

Abstract--The extraction of geological lineaments from digital satellite data is a fundamental application in remote sensing. The location of geological lineaments such as faults and dykes are of interest for a range of applications, particularly because of their association with hydrothermal mineralization. Although a wide range of applications have utilized computer vision techniques, a standard workflow for application of these techniques to mineral exploration is lacking. We present a framework for extracting geological lineaments using computer vision techniques which is a combination of edge detection and line extraction algorithms for extracting geological lineaments using optical remote sensing data. It features ancillary computer vision techniques for reducing data dimensionality, removing noise and enhancing the expression of lineaments. We test the proposed framework on Landsat 8 data of a mineral-rich portion of the Gascoyne Province in Western Australia using different dimension reduction techniques and convolutional filters. To validate the results, the extracted lineaments are compared to our manual photointerpretation and geologically mapped structures by the Geological Survey of Western Australia (GSWA). The results show that the best correlation between our extracted geological lineaments and the GSWA geological lineament map is achieved by applying a minimum noise fraction transformation and a Laplacian filter. Application of a directional filter instead shows a stronger correlation with the output of our manual photointerpretation and known sites of hydrothermal mineralization. Hence, our framework using either filter can be used for mineral prospectivity mapping in other regions where faults are exposed and observable in optical remote sensing data. IGITAL satellite data with different spatial and spectral resolution are available for almost every locality on the Earth's land surface [1]-[5]. This enables the procurement of detailed information from surficial features and processes at different scales. Linear features are considered as one of the most important surficial features in different fields of study [6]-[8]. R. Scalzo is with the Centre for Translational Data Science, University of Sydney, Sydney, NSW 2006, Australia (email: richard.scalzo@sydney.edu.au). Linear features represent the expression of some degree of linearity of a single or diverse grouping of both natural and cultural features [9], [10].

Bank of New Zealand is ramping up cross-channel fraud prevention for its customers through a multi-million dollar deal with IBM. The banking giant said its aim was to provide its customers with the ability to bank securely while delivering a positive customer experience. Furthermore, BNZ said legacy systems were designed to see and stop easily recognisable fraud patterns, however, modern "anytime, anywhere" banking on mobile devices has made fraud detection much more challenging. "Banks' time to respond is also shrinking as real-time payments mean there are just milliseconds to detect and prevent theft before it's too late," the bank said. Global card fraud losses are also increasing.

Dark pools are electronic trading platforms that have emerged in the past decade in advanced markets. They allow traders to buy or sell large blocks of shares without having to disclose their identities, the volumes or prices, unlike traditional exchanges. They are popular with asset-management companies, pension funds and insurance firms which need to conduct a lot of large transactions, because they are cheaper and easier to carry out via electronic trading platforms. Merrin founded Liquidnet in 2001 in the US and later expanded into Europe and Asia-Pacific. The platform has seen trading volume in Asia-Pacific of US$42 billion so far this year, up 57 per cent from a year earlier.

Professor Martin Carroll is an Executive General Manager, Academic & Provost Manukau Institute of Technology, New Zealand. In 2017, Martin co-founded with Blackboard, a global forum for the Ethical Use of AI in higher education. Drawing upon the projections of realists and the visions of sci-fi, Martin's entertaining and thought provoking talk will leave you viewing education in an entirely new light! One clear conclusion is the need for ethical guidance for AI and AI-type technologies. Drawing upon the projections of realists and the visions of sci-fi, Martin's entertaining and thought-provoking talk will leave you viewing education in an entirely new light.

Artificial intelligence, extended reality and blockchain are some of the most talked about emerging digital technologies. While not typically associated with the healthcare industry, these technologies are having a big impact on the supply chain of pharmaceutical products. To meet the expectations of today's patients, supply chains need to be smart, connected and agile so they provide an efficient and personalised patient experience. Companies that don't implement digital technologies put themselves at risk of being unable to do the one thing that matters most: better serve their patients. The Therapeutic Goods Administration recently said logistical difficulties, recalls, manufacturing issues and unexpected increases in demand were the main causes of shortages between January and August this year, highlighting the need for global companies operating in Australia to rethink how they can make their supply chains more efficient.

Earlier this year (May 2018) Microsoft announced ML.NET, an open source and cross-platform machine learning framework built for .NET developers. It is exciting news to be able to integrate custom machine learning with .NET/C# applications. Although ML.NET is still in preview release version 0.5.0 at the time of writing, you can test drive it to explore the potential power of the framework. There are already a number of tutorials for ML.NET available from Microsoft and third parties. However, the example data sources are mostly flat files in the format of TSV (Tab Separated Values).

One of the key targets Enlitic is focused on is early detection of lung cancer by combining biopsies along with existing medical data to be able to diagnose lung cancer earlier than with the traditional medical methods. While based in San Francisco, Enlitic is present in Japan, Canada, Australia, and China. AI is the buzzword startups are very keen on using when describing their products. We've been seeing ideas on what it could do in movies for decades. So what qualifies as AI?

Reinforcement learning has shown great potential in generalizing over raw sensory data using only a single neural network for value optimization. There are several challenges in the current state-of-the-art reinforcement learning algorithms that prevent them from converging towards the global optima. It is likely that the solution to these problems lies in short- and long-term planning, exploration and memory management for reinforcement learning algorithms. Games are often used to benchmark reinforcement learning algorithms as they provide a flexible, reproducible, and easy to control environment. Regardless, few games feature a state-space where results in exploration, memory, and planning are easily perceived. This paper presents The Dreaming Variational Autoencoder (DVAE), a neural network based generative modeling architecture for exploration in environments with sparse feedback. We further present Deep Maze, a novel and flexible maze engine that challenges DVAE in partial and fully-observable state-spaces, long-horizon tasks, and deterministic and stochastic problems. We show initial findings and encourage further work in reinforcement learning driven by generative exploration.

Software defined networking (SDN) represents a promising networking architecture that combines central management and network programmability. SDN separates the control plane from the data plane and moves the network management to a central point, called the controller, that can be programmed and used as the brain of the network. Recently, the research community has showed an increased tendency to benefit from the recent advancements in the artificial intelligence (AI) field to provide learning abilities and better decision making in SDN. In this study, we provide a detailed overview of the recent efforts to include AI in SDN. Our study showed that the research efforts focused on three main sub-fields of AI namely: machine learning, meta-heuristics and fuzzy inference systems. Accordingly, in this work we investigate their different application areas and potential use, as well as the improvements achieved by including AI-based techniques in the SDN paradigm.