Melbourne man Milorad "Michael" Trkulja has won his high court battle to sue the search engine Google for defamation over images and search results that link him to the Melbourne criminal underworld. Trkulja said he would continue legal action against Google until it removed his name and photos from the internet. Trkulja, who was shot in the back in a Melbourne restaurant in 2004, successfully argued in the Victorian supreme court in 2012 that Google defamed him by publishing photos of him linked to hardened criminals of Melbourne's underworld. Four years later the Victorian court of appeal overturned the decision, finding the case had no prospect of successfully proving defamation. The high court disputed that ruling in a judgment on Wednesday and ordered Google to pay Trkulja's legal costs.

We propose a simple and easy to implement neural network compression algorithm that achieves results competitive with more complicated state-of-the-art methods. The key idea is to modify the original optimization problem by adding K independent Gaussian priors (corresponding to the k-means objective) over the network parameters to achieve parameter quantization, as well as an L1 penalty to achieve pruning. Unlike many existing quantization-based methods, our method uses hard clustering assignments of network parameters, which adds minimal change or overhead to standard network training. We also demonstrate experimentally that tying neural network parameters provides less gain in generalization performance than changing network architecture and connectivity patterns entirely.

The 2017 Fake News Challenge Stage 1 (FNC-1) shared task addressed a stance classification task as a crucial first step towards detecting fake news. To date, there is no in-depth analysis paper to critically discuss FNC-1's experimental setup, reproduce the results, and draw conclusions for next-generation stance classification methods. In this paper, we provide such an in-depth analysis for the three top-performing systems. We first find that FNC-1's proposed evaluation metric favors the majority class, which can be easily classified, and thus overestimates the true discriminative power of the methods. Therefore, we propose a new F1-based metric yielding a changed system ranking. Next, we compare the features and architectures used, which leads to a novel feature-rich stacked LSTM model that performs on par with the best systems, but is superior in predicting minority classes. To understand the methods' ability to generalize, we derive a new dataset and perform both in-domain and cross-domain experiments. Our qualitative and quantitative study helps interpreting the original FNC-1 scores and understand which features help improving performance and why. Our new dataset and all source code used during the reproduction study are publicly available for future research.

Google Home speakers can already perform two commands at the same time. But what if you live in a particularly connected household, where you may need to juggle more tasks? The company has an answer: support three simultaneous requests. So long as you form full queries with "and" in between, you can multitask like a pro using only the spoken word. This could be particularly helpful if you want to turn on the lights, increase the temperature and play some tunes without having a Routine in place.

Melbourne: A fleet of cars saddled with 3D mapping equipment and mounted with spatial technology are driving thousands of kilometres of Australian roads, as part of a trial of a Satellite-Based Augmentation System (SBAS), which will deliver map accuracy to 10 centimetres and support the operations of autonomous vehicles. The trial, undertaken by HERE Technologies in partnership with the Australia and New Zealand Cooperative Research Centre for Spatial Information (CRCSI), is enabling the company to challenge conventional mapping processes and ensure the delivery of a high-definition (HD) map to support Highly Automated Driving (HAD), advance road vehicle safety and warnings, and improve traffic flow management in Australia and New Zealand. CRCSI CEO, Dr Graeme Kernich said SBAS technology has potential uses in a range of sectors, not least in ensuring Australia has the necessary infrastructure required for autonomous vehicles. "This trial is one important piece of the puzzle to enable a safe, efficient and sustainable future for Australia in an autonomous world," said Dr Kernich. "By pooling resources business, and government can create the infrastructure needed to support the transportation systems and cities of the future."

If you fancy a flutter on the World Cup, you may want to put your money on Brazil. According to more than a million simulations of the tournament, the five-time winners are the favourites to lift the coveted trophy on July 15th. The study, from experts at Goldman Sachs, also predicts England will make it to the quarterfinal stage where they will be knocked out by Germany. The financial firm made the predictions after feeding an AI data on team strategy, the strengths and weaknesses of individual players, and recent team results. Brazil are set to be crowned this year's Fifa World Cup winners, according to a prediction driven by more than one million simulations of the tournament.

Stochastic zeroth-order (SZO), or gradient-free, optimization allows to optimize arbitrary functions by relying only on function evaluations under parameter perturbations, however, the iteration complexity of SZO methods suffers a factor proportional to the dimensionality of the perturbed function. We show that in scenarios with natural sparsity patterns as in structured prediction applications, this factor can be reduced to the expected number of active features over input-output pairs. We give a general proof that applies sparse SZO optimization to Lipschitz-continuous, nonconvex, stochastic objectives, and present an experimental evaluation on linear bandit structured prediction tasks with sparse word-based feature representations that confirm our theoretical results.

This dissertation investigates the use of one-sided classification algorithms in the application of separating hazardous chlorinated solvents from other materials, based on their Raman spectra. The experimentation is carried out using a new one-sided classification toolkit that was designed and developed from the ground up. In the one-sided classification paradigm, the objective is to separate elements of the target class from all outliers. These one-sided classifiers are generally chosen, in practice, when there is a deficiency of some sort in the training examples. Sometimes outlier examples can be rare, expensive to label, or even entirely absent. However, this author would like to note that they can be equally applicable when outlier examples are plentiful but nonetheless not statistically representative of the complete outlier concept. It is this scenario that is explicitly dealt with in this research work. In these circumstances, one-sided classifiers have been found to be more robust that conventional multi-class classifiers. The term "unexpected" outliers is introduced to represent outlier examples, encountered in the test set, that have been taken from a different distribution to the training set examples. These are examples that are a result of an inadequate representation of all possible outliers in the training set. It can often be impossible to fully characterise outlier examples given the fact that they can represent the immeasurable quantity of "everything else" that is not a target. The findings from this research have shown the potential drawbacks of using conventional multi-class classification algorithms when the test data come from a completely different distribution to that of the training samples.

We show empirically that the optimal strategy of parameter averaging in a minmax convex-concave game setting is also strikingly effective in the non convex-concave GAN setting, specifically alleviating the convergence issues associated with cycling behavior observed in GANs. We show that averaging over generator parameters outside of the trainig loop consistently improves inception and FID scores on different architectures and for different GAN objectives. We provide comprehensive experimental results across a range of datasets, bilinear games, mixture of Gaussians, CIFAR-10, STL-10, CelebA and ImageNet, to demonstrate its effectiveness. We achieve state-of-the-art results on CIFAR-10 and produce clean CelebA face images, demonstrating that averaging is one of the most effective techniques for training highly performant GANs.

Deep learning has enabled traditional reinforcement learning methods to deal with high-dimensional problems. However, one of the disadvantages of deep reinforcement learning methods is the limited exploration capacity of learning agents. In this paper, we introduce an approach that integrates human strategies to increase the exploration capacity of multiple deep reinforcement learning agents. We also report the development of our own multi-agent environment called Multiple Tank Defence to simulate the proposed approach. The results show the significant performance improvement of multiple agents that have learned cooperatively with human strategies. This implies that there is a critical need for human intellect teamed with machines to solve complex problems. In addition, the success of this simulation indicates that our developed multi-agent environment can be used as a testbed platform to develop and validate other multi-agent control algorithms. Details of the environment implementation can be referred to http://www.deakin.edu.au/~thanhthi/madrl_human.htm