Algorithms'learn' from the data they are trained on to create a machine learning model that can perform a given task effectively without needing specific instructions, such as making predictions or accurately classifying images and emails. These techniques are already used widely, for example to identify spam emails, diagnose diseases from X-rays, predict crop yields and will soon drive our cars. While the technology holds enormous potential to positively transform our world, artificial intelligence and machine learning are vulnerable to adversarial attacks, a technique employed to fool machine learning models through the input of malicious data causing them to malfunction. Dr Richard Nock, machine learning group leader at CSIRO's Data61 said that by adding a layer of noise (i.e. an adversary) over an image, attackers can deceive machine learning models into misclassifying the image. "Adversarial attacks have proven capable of tricking a machine learning model into incorrectly labelling a traffic stop sign as speed sign, which could have disastrous effects in the real world. "Our new techniques prevent adversarial attacks using a process similar to vaccination," Dr Nock said. "We implement a weak version of an adversary, such as small modifications or distortion to a collection of images, to create a more'difficult' training data set.

Business today is more than simply matching traditional competitors, it's about exploiting digital technologies to create new opportunities, and being able to repeat this. The economy is quickly going digital and Australian businesses must evolve into Modern Digital Businesses (MDBs) which strategically use intelligence assets to improve operations and deploy new products and services, in order to stay competitive and create value for their customers. A group of digital business leaders recently gathered at ThoughtWorks Live in Sydney and Melbourne, to share their insights into how organisations can take advantage of data to adapt and thrive in the digital economy. This report includes strategic and practical advice taken from the event for any business leader – regardless of their organisation's digital maturity – on best practices for taking advantage of data and driving change. A Continuous Intelligence (CI) framework starts with the process of acquiring data and, with the help of analytics and machine learning, derive insights from it to be able to make confident decisions and actions – which are in turn reviewed and validated, to ensure the organisation continuously improves its decision-making capabilities. Steps organisations can take to apply CI to building an MDB, which is agile and technology-driven are also covered.

To train this AI, we only need to input articles, books, languages, and internet sources. Therefore, the bias and complexity could be reduced and minimized. The model could be used for general purpose selection. Apart from that, the existing voice recognizers and chatbots can support better candidate communication. The aim of first calls by contingent firms is simply laying out the basic job descriptions and asking for a YES or NO, which chatbots could easily handle.

The myth that modern wheat varieties are more heavily reliant on pesticides and fertilisers than older varieties has been debunked by new research. The University of Queensland's Dr Kai Voss-Fels said modern wheat varieties have out-performed older varieties in side-by-side field trials under both optimum and harsh growing conditions. "There is a view that intensive selection and breeding, which has produced the high-yielding wheat cultivars used in modern cropping, has also made them less resilient and more dependent on chemicals to thrive," Dr Voss-Fels said. "However, the data published today unequivocally shows that modern wheat out-performs older varieties, even under conditions of reduced amounts of fertilisers, fungicides and water. "We also found that genetic diversity within the relatively narrow modern wheat gene pool is rich enough to potentially generate a further 23 per cent increase in yields."

We propose behavior-driven optimization via Wasserstein distances (WDs) to improve several classes of state-of-the-art reinforcement learning (RL) algorithms. We show that WD regularizers acting on appropriate policy embeddings efficiently incorporate behavioral characteristics into policy optimization. We demonstrate that they improve Evolution Strategy methods by encouraging more efficient exploration, can be applied in imitation learning and to speed up training of Trust Region Policy Optimization methods. Since the exact computation of WDs is expensive, we develop approximate algorithms based on the combination of different methods: dual formulation of the optimal transport problem, alternating optimization and random feature maps, to effectively replace exact WD computations in the RL tasks considered. We provide theoretical analysis of our algorithms and exhaustive empirical evaluation in a variety of RL settings.

Imitation Learning describes the problem of recovering an expert policy from demonstrations. While inverse reinforcement learning approaches are known to be very sample-efficient in terms of expert demonstrations, they usually require problem-dependent reward functions or a (task-)specific reward-function regularization. In this paper, we show a natural connection between inverse reinforcement learning approaches and Optimal Transport, that enables more general reward functions with desirable properties (e.g., smoothness). Based on our observation, we propose a novel approach called Wasserstein Adversarial Imitation Learning. Our approach considers the Kantorovich potentials as a reward function and further leverages regularized optimal transport to enable large-scale applications. In several robotic experiments, our approach outperforms the baselines in terms of average cumulative rewards and shows a significant improvement in sample-efficiency, by requiring just one expert demonstration.

In this work we present a novel approach to solving concurrent multiagent planning problems in which several agents act in parallel. Our approach relies on a compilation from concurrent multiagent planning to classical planning, allowing us to use an off-the-shelf classical planner to solve the original multiagent problem. The solution can be directly interpreted as a concurrent plan that satisfies a given set of concurrency constraints, while avoiding the exponential blowup associated with concurrent actions. Our planner is the first to handle action effects that are conditional on what other agents are doing. Theoretically, we show that the compilation is sound and complete. Empirically, we show that our compilation can solve challenging multiagent planning problems that require concurrent actions.

Following the attacks in Christchurch, New Zealand, in March, social media companies have once again come under growing pressure to "do …

Artificial intelligence is one of the technologies with the most transformative potential in business. According to research by McKinsey, 70 per cent of companies are likely to have adopted at least one form of AI by 2030. This will contribute to an additional $13tr of global economic activity. Machine learning – a subset of artificial intelligence – enables machines to get better at executing tasks without human intervention, by finding patterns in data, and learning from their experience. It's no surprise, therefore, that there has been an explosion in the number of machine learning companies worldwide.

In multiagent learning, agents interact in inherently nonstationary environments due to their concurrent policy updates. It is, therefore, paramount to develop and analyze algorithms that learn effectively despite these nonstationarities. A number of works have successfully conducted this analysis under the lens of evolutionary game theory (EGT), wherein a population of individuals interact and evolve based on biologically-inspired operators. These studies have mainly focused on establishing connections to value-iteration based approaches in stateless or tabular games. We extend this line of inquiry to formally establish links between EGT and policy gradient (PG) methods, which have been extensively applied in single and multiagent learning. We pinpoint weaknesses of the commonly-used softmax PG algorithm in adversarial and nonstationary settings and contrast PG's behavior to that predicted by replicator dynamics (RD), a central model in EGT. We consequently provide theoretical results that establish links between EGT and PG methods, then derive Neural Replicator Dynamics (NeuRD), a parameterized version of RD that constitutes a novel method with several advantages. First, as NeuRD reduces to the well-studied no-regret Hedge algorithm in the tabular setting, it inherits no-regret guarantees that enable convergence to equilibria in games. Second, NeuRD is shown to be more adaptive to nonstationarity, in comparison to PG, when learning in canonical games and imperfect information benchmarks including Poker. Thirdly, modifying any PG-based algorithm to use the NeuRD update rule is straightforward and incurs no added computational costs. Finally, while single-agent learning is not the main focus of the paper, we verify empirically that NeuRD is competitive in these settings with a recent baseline algorithm.