While use of artificial intelligence and machine learning is high amongst IT decision makers, more than half are unsure what the technology actually means, new research has revealed. Webroot has released its annual artificial intelligence and machine learning report, which reveals how IT professionals perceive and use these advancing technologies in business. According to the research, while 93% of IT decision makers in Australia and New Zealand use artificial intelligence and machine learning, more than half (51%) admit they are unsure what the technology means. Although understanding around these tools is increasing (64% of global respondents were unsure what artificial intelligence and machine learning meant in 2020), this is happening at a significantly slower pace than the adoption rate. Webroot says this lack of understanding may be why Australia was the country where enterprises were most likely to cite struggling to keep up with the latest technology as the reason they were unable to prevent a cyberattack in the last year.

Neural network robustness has become a central topic in machine learning in recent years. Most training algorithms that improve the model's robustness to adversarial and common corruptions also introduce a large computational overhead, requiring as many as ten times the number of forward and backward passes in order to converge. To combat this inefficiency, we propose BulletTrain $-$ a boundary example mining technique to drastically reduce the computational cost of robust training. Our key observation is that only a small fraction of examples are beneficial for improving robustness. BulletTrain dynamically predicts these important examples and optimizes robust training algorithms to focus on the important examples. We apply our technique to several existing robust training algorithms and achieve a 2.1$\times$ speed-up for TRADES and MART on CIFAR-10 and a 1.7$\times$ speed-up for AugMix on CIFAR-10-C and CIFAR-100-C without any reduction in clean and robust accuracy.

Despite Graph Neural Networks (GNNs) have achieved remarkable accuracy, whether the results are trustworthy is still unexplored. Previous studies suggest that many modern neural networks are over-confident on the predictions, however, surprisingly, we discover that GNNs are primarily in the opposite direction, i.e., GNNs are under-confident. Therefore, the confidence calibration for GNNs is highly desired. In this paper, we propose a novel trustworthy GNN model by designing a topology-aware post-hoc calibration function. Specifically, we first verify that the confidence distribution in a graph has homophily property, and this finding inspires us to design a calibration GNN model (CaGCN) to learn the calibration function. CaGCN is able to obtain a unique transformation from logits of GNNs to the calibrated confidence for each node, meanwhile, such transformation is able to preserve the order between classes, satisfying the accuracy-preserving property. Moreover, we apply the calibration GNN to self-training framework, showing that more trustworthy pseudo labels can be obtained with the calibrated confidence and further improve the performance. Extensive experiments demonstrate the effectiveness of our proposed model in terms of both calibration and accuracy.

Accurate inference of fine-grained traffic flow from coarse-grained one is an emerging yet crucial problem, which can help greatly reduce the number of traffic monitoring sensors for cost savings. In this work, we notice that traffic flow has a high correlation with road network, which was either completely ignored or simply treated as an external factor in previous works. To facilitate this problem, we propose a novel Road-Aware Traffic Flow Magnifier (RATFM) that explicitly exploits the prior knowledge of road networks to fully learn the road-aware spatial distribution of fine-grained traffic flow. Specifically, a multi-directional 1D convolutional layer is first introduced to extract the semantic feature of the road network. Subsequently, we incorporate the road network feature and coarse-grained flow feature to regularize the short-range spatial distribution modeling of road-relative traffic flow. Furthermore, we take the road network feature as a query to capture the long-range spatial distribution of traffic flow with a transformer architecture. Benefiting from the road-aware inference mechanism, our method can generate high-quality fine-grained traffic flow maps. Extensive experiments on three real-world datasets show that the proposed RATFM outperforms state-of-the-art models under various scenarios.

Double Q-learning is a classical method for reducing overestimation bias, which is caused by taking maximum estimated values in the Bellman operation. Its variants in the deep Q-learning paradigm have shown great promise in producing reliable value prediction and improving learning performance. However, as shown by prior work, double Q-learning is not fully unbiased and suffers from underestimation bias. In this paper, we show that such underestimation bias may lead to multiple non-optimal fixed points under an approximated Bellman operator. To address the concerns of converging to non-optimal stationary solutions, we propose a simple but effective approach as a partial fix for the underestimation bias in double Q-learning. This approach leverages an approximate dynamic programming to bound the target value. We extensively evaluate our proposed method in the Atari benchmark tasks and demonstrate its significant improvement over baseline algorithms.

According to the research, while 93% of IT decision makers in Australia and New Zealand use artificial intelligence and machine learning, more than half (51%) …

Can artificial intelligence help us understand transmissible infections better? As people the world over become accepting of testing regimes required to diagnose, monitor, and assess outbreaks of the COVID-19, researchers from the University of Southern Queensland are attempting to change the way testing for other types of transmissible infections are understood and utilised – using Artificial Intelligence (AI). A cross-disciplinary research team made up of four USQ experts and an expert colleague from The University of Queensland have been awarded a $500,000 competitive research grant from the Australian Government as part of its First National Blood Borne Viruses and Sexually Transmissible Infections Research Strategy 2021 – 2025. The grant will support the development of a mobile app, supported by AI, centred around sexual health risk behaviours, and screening and testing for Sexually Transmissible Infections, or STIs. Dr Zhaohui Tang and Professor Yan Li from USQ's School of Sciences will lead the project.

At the Goodwood festival of speed near Chichester, the crowds gathered at the hill-climb circuit to watch the world's fastest cars roar past, as they do every year. But not far from the high-octane action, there was a new, and quieter, attraction: a display of the latest electric vehicles, from the £28,000 Mini Electric to the £2m Lotus Evija hypercar. Even here, at one of the biggest events in Britain's petrolhead calendar, it's clear the days of the internal combustion engine are numbered. As countries strive to meet stringent carbon-emission targets, and vehicle-makers phase out combustion engines, 145m electric vehicles are predicted to be on the roads within a decade, up from 11m last year. The car batteries they require, along with storage batteries for solar and wind power, have sent demand for metals soaring, taking mining firms to the bottom of the sea in the hunt for those metals.

There has been increasing interest in characterising the error behaviour of systems which contain deep learning models before deploying them into any safety-critical scenario. However, characterising such behaviour usually requires large-scale testing of the model that can be extremely computationally expensive for complex real-world tasks. For example, tasks involving compute intensive object detectors as one of their components. In this work, we propose an approach that enables efficient large-scale testing using simplified low-fidelity simulators and without the computational cost of executing expensive deep learning models. Our approach relies on designing an efficient surrogate model corresponding to the compute intensive components of the task under test. We demonstrate the efficacy of our methodology by evaluating the performance of an autonomous driving task in the Carla simulator with reduced computational expense by training efficient surrogate models for PIXOR and CenterPoint LiDAR detectors, whilst demonstrating that the accuracy of the simulation is maintained.

In collaborative filtering, it is an important way to make full use of social information to improve the recommendation quality, which has been proved to be effective because user behavior will be affected by her friends. However, existing works leverage the social relationship to aggregate user features from friends' historical behavior sequences in a user-level indirect paradigm. A significant defect of the indirect paradigm is that it ignores the temporal relationships between behavior events across users. In this paper, we propose a novel time-aware sequential recommendation framework called Social Temporal Excitation Networks (STEN), which introduces temporal point processes to model the fine-grained impact of friends' behaviors on the user s dynamic interests in an event-level direct paradigm. Moreover, we propose to decompose the temporal effect in sequential recommendation into social mutual temporal effect and ego temporal effect. Specifically, we employ a social heterogeneous graph embedding layer to refine user representation via structural information. To enhance temporal information propagation, STEN directly extracts the fine-grained temporal mutual influence of friends' behaviors through the mutually exciting temporal network. Besides, the user s dynamic interests are captured through the self-exciting temporal network. Extensive experiments on three real-world datasets show that STEN outperforms state-of-the-art baseline methods. Moreover, STEN provides event-level recommendation explainability, which is also illustrated experimentally.