Developing a master plan to transform John F. Kennedy International Airport in New York. Keeping beachgoers safe from polluted waters in New Zealand with advanced analytics. Those are just a few of the thousands of complex projects delivered each year by Mott MacDonald, a global engineering, management and development consulting firm headquartered in London. With 180 principal offices in 50 countries, the company helps solve some of the world's most urgent social, environmental and economic challenges. Because Mott MacDonald doesn't create physical products, its success relies on the knowledge and expertise of its 16,000 employees.

Artificial Intelligence in an open banking environment has revolutionised the way customers approach and leverage their financial information. The face of banking is changing across Australia and around the globe. With the launch of open banking and the Australian Consumer Data Right, the latest digital technologies and an ever-changing regulatory landscape, customers are demanding enhanced experiences and increased value for money. Artificial Intelligence (AI) is just one technology that is enabling banks and financiers to provide increased efficiencies behind the scenes, new product lines upfront, and to ensure a heightened customer experience for all. The customer experience has become the significant driver of any brand's growth and often their point of differentiation.

The implications of climate change are considered by many to be the greatest crisis confronting us globally. Most focus on the doomsday scenarios of melting glaciers, rising flood waters engulfing coastal cities, and species becoming extinct. The other challenge is feeding the world's growing populace. There are currently about 7.7 billion people living on Earth, and that number is projected to increase to almost 10 billion by 2050. Estimates put the number of those people who are chronically hungry at almost 1 billion even now.

Autonomous vehicles might be a double-edged sword for the driving sector, and they're well on the way - people with personal interest across the board are already sponsoring legislation to speed up the process of getting them onto our roads. With well-developed local autonomous driving laws already taking shape across various parts of Australia, law-makers are on track to see to a smooth transition that takes everyone on board. There are a number of predicted outcomes for how autonomous vehicles could impact real human drivers. In the worst-case scenario, autonomous vehicles might drain away jobs, taking millions from the sector. In a zero-sum scenario, there might be a decrease in certain types of driving-related jobs and an increase in others. And in the best case, we'll see an increase in desirable jobs and a reduction in more physically demanding jobs in the sector.

Pramod, 40, a software engineer with a large IT company, is having a mid-life crisis. Bored of writing hundreds of lines of code every day, he wants to quit the monotonous job and shift to a more challenging role that also gives him a good package. However, he is clueless on the right skill sets he needs to acquire. Pramod is not alone in this predicament. There are thousands of such mid-level software professionals who either want to make a course correction or have been told by their organisation to acquire new skills if they are to remain relevant.

Artificial intelligence should respect human rights, diversity and privacy -- while being a far cry from Terminator-style robots -- according to new federal ethics guidelines. Technology Minister Karen Andrews will today release an eight-point guidance she wants companies to adopt in a bid to prevent people from being exploited. The guidelines stipulate all AI should benefit individuals, society and the environment. It should prevent discrimination, respect privacy and only operate in accordance with their intended purpose. The guidelines also recommend human oversight of AI always be enabled and there should be timely processes to allow people to challenge the use or output of information.

Some of the biggest businesses in Australia will trial a series of eight principles around artificial intelligence, developed as part of the Morrison Government's AI Ethics Framework. NAB, Commonwealth Bank, Telstra, Microsoft and Flamingo AI have signed up to test the principles to ensure they deliver practical benefits and translate into real world solutions. Minister for Industry, Science and Technology Karen Andrews said AI is a powerful technology that can create jobs, boost the economy and improve our quality of life and is an important part of the Government's economic plan. "The Morrison Government is determined to create an environment where AI helps the economy and everyday Australians to thrive. The eight AI ethics principles are just one part of this vision," Minister Andrews said.

With the increasing complexity of modern power systems, conventional dynamic load modeling with ZIP and induction motors (ZIP + IM) is no longer adequate to address the current load characteristic transitions. In recent years, the WECC composite load model (WECC CLM) has shown to effectively capture the dynamic load responses over traditional load models in various stability studies and contingency analyses. However, a detailed WECC CLM model typically has a high degree of complexity, with over one hundred parameters, and no systematic approach to identifying and calibrating these parameters. Enabled by the wide deployment of PMUs and advanced deep learning algorithms, proposed here is a double deep Q-learning network (DDQN)-based, two-stage load modeling framework for the WECC CLM. This two-stage method decomposes the complicated WECC CLM for more efficient identification and does not require explicit model details. In the first stage, the DDQN agent determines an accurate load composition. In the second stage, the parameters of the WECC CLM are selected from a group of Monte-Carlo simulations. The set of selected load parameters is expected to best approximate the true transient responses. The proposed framework is verified using an IEEE 39-bus test system on commercial simulation platforms.

Support Vector Machines (SVMs) can solve structured multi-output learning problems such as multi-label classification, multiclass classification and vector regression. SVM training is expensive especially for large and high dimensional datasets. The bottleneck of the SVM training often lies in the kernel value computation. In many real-world problems, the same kernel values are used in many iterations during the training, which makes the caching of kernel values potentially useful. The majority of the existing studies simply adopt the LRU (least recently used) replacement strategy for caching kernel values. However, as we analyze in this paper, the LRU strategy generally achieves high hit ratio near the final stage of the training, but does not work well in the whole training process. Therefore, we propose a new caching strategy called EFU (less frequently used) which replaces the less frequently used kernel values that enhances LFU (least frequently used). Our experimental results show that EFU often has 20\% higher hit ratio than LRU in the training with the Gaussian kernel. To further optimize the strategy, we propose a caching strategy called HCST (hybrid caching for the SVM training), which has a novel mechanism to automatically adapt the better caching strategy in the different stages of the training. We have integrated the caching strategy into ThunderSVM, a recent SVM library on many-core processors. Our experiments show that HCST adaptively achieves high hit ratios with little runtime overhead among different problems including multi-label classification, multiclass classification and regression problems. Compared with other existing caching strategies, HCST achieves 20\% more reduction in training time on average.

Vision based localization is the problem of inferring the pose of the camera given a single image. One solution to this problem is to learn a deep neural network to infer the pose of a query image after learning on a dataset of images with known poses. Another more commonly used approach rely on image retrieval where the query image is compared against the database of images and its pose is inferred with the help of the retrieved images. The latter approach assumes that images taken from the same places consists of the same landmarks and, thus would have similar feature representations. These representation can be learned using full supervision to be robust to different variations in capture conditions like time of the day and weather. In this work, we introduce a framework to enhance the performance of these retrieval based localization methods by taking into account the additional information including GPS coordinates and temporal neighbourhood of the images provided by the acquisition process in addition to the descriptor similarity of pairs of images in the reference or query database which is used traditionally for localization. Our method constructs a graph based on this additional information and use it for robust retrieval by smoothing the feature representation of reference and/or query images. We show that the proposed method is able to significantly improve the localization accuracy on two large scale datasets over the baselines.