"The big tech is banking heavily on AI, Cloud and 5G technologies to retain customers and drive growth" A global emergency can smother your business, government lawsuits can break your company, competitors with trillion-dollar market value can wipe your organisation off the map. But what would happen when all three come together in the same year? The pandemic brought the world to a standstill. The internet giants, however, came out of it unscathed. Apple, Amazon, Google and Facebook, popularly known as the big four, have not only survived a combination of calamities but registered profits and left the Wall Street analysts dumbfounded.

At the peak of the Covid-19 pandemic in 2020, Australian transport agency Transport for New South Wales (NSW) had to restore public confidence in the state's transportation network and curb the spread of the disease. One of the ways it did that was to analyse the travel history recorded by Opal transit cards – with an individual's permission – and inform the commuter if the regular buses and train services that they had been taking were Covid-safe. Chris Bennetts, executive director for digital product delivery at Transport for NSW, said those insights were derived using a machine learning model that predicts how full a bus or train carriage was going to be at a given time. Based on the predictions, commuters would be advised if they could continue using their regular services or switch to a different service or mode of transport. "That was interesting for us because it was our first foray into personalisation to offer more choices for customers," said Bennetts.

Because of the superior feature representation ability of deep learning, various deep Click-Through Rate (CTR) models are deployed in the commercial systems by industrial companies. To achieve better performance, it is necessary to train the deep CTR models on huge volume of training data efficiently, which makes speeding up the training process an essential problem. Different from the models with dense training data, the training data for CTR models is usually high-dimensional and sparse. To transform the high-dimensional sparse input into low-dimensional dense real-value vectors, almost all deep CTR models adopt the embedding layer, which easily reaches hundreds of GB or even TB. Since a single GPU cannot afford to accommodate all the embedding parameters, when performing distributed training, it is not reasonable to conduct the data-parallelism only. Therefore, existing distributed training platforms for recommendation adopt model-parallelism. Specifically, they use CPU (Host) memory of servers to maintain and update the embedding parameters and utilize GPU worker to conduct forward and backward computations. Unfortunately, these platforms suffer from two bottlenecks: (1) the latency of pull \& push operations between Host and GPU; (2) parameters update and synchronization in the CPU servers. To address such bottlenecks, in this paper, we propose the ScaleFreeCTR: a MixCache-based distributed training system for CTR models. Specifically, in SFCTR, we also store huge embedding table in CPU but utilize GPU instead of CPU to conduct embedding synchronization efficiently. To reduce the latency of data transfer between both GPU-Host and GPU-GPU, the MixCache mechanism and Virtual Sparse Id operation are proposed. Comprehensive experiments and ablation studies are conducted to demonstrate the effectiveness and efficiency of SFCTR.

Rationalizing which parts of a molecule drive the predictions of a molecular graph convolutional neural network (GCNN) can be difficult. To help, we propose two simple regularization techniques to apply during the training of GCNNs: Batch Representation Orthonormalization (BRO) and Gini regularization. BRO, inspired by molecular orbital theory, encourages graph convolution operations to generate orthonormal node embeddings. Gini regularization is applied to the weights of the output layer and constrains the number of dimensions the model can use to make predictions. We show that Gini and BRO regularization can improve the accuracy of state-of-the-art GCNN attribution methods on artificial benchmark datasets. In a real-world setting, we demonstrate that medicinal chemists significantly prefer explanations extracted from regularized models. While we only study these regularizers in the context of GCNNs, both can be applied to other types of neural networks

Marginal-likelihood based model-selection, even though promising, is rarely used in deep learning due to estimation difficulties. Instead, most approaches rely on validation data, which may not be readily available. In this work, we present a scalable marginal-likelihood estimation method to select both the hyperparameters and network architecture based on the training data alone. Some hyperparameters can be estimated online during training, simplifying the procedure. Our marginal-likelihood estimate is based on Laplace's method and Gauss-Newton approximations to the Hessian, and it outperforms cross-validation and manual-tuning on standard regression and image classification datasets, especially in terms of calibration and out-of-distribution detection. Our work shows that marginal likelihoods can improve generalization and be useful when validation data is unavailable (e.g., in nonstationary settings).

Neuro-Symbolic Artificial Intelligence -- the combination of symbolic methods with methods that are based on artificial neural networks -- has a long-standing history. In this article, we provide a structured overview of current trends, by means of categorizing recent publications from key conferences. The article is meant to serve as a convenient starting point for research on the general topic.

Signed languages are the primary means of communication for many deaf and hard of hearing individuals. Since signed languages exhibit all the fundamental linguistic properties of natural language, we believe that tools and theories of Natural Language Processing (NLP) are crucial towards its modeling. However, existing research in Sign Language Processing (SLP) seldom attempt to explore and leverage the linguistic organization of signed languages. This position paper calls on the NLP community to include signed languages as a research area with high social and scientific impact. We first discuss the linguistic properties of signed languages to consider during their modeling. Then, we review the limitations of current SLP models and identify the open challenges to extend NLP to signed languages. Finally, we urge (1) the adoption of an efficient tokenization method; (2) the development of linguistically-informed models; (3) the collection of real-world signed language data; (4) the inclusion of local signed language communities as an active and leading voice in the direction of research.

Over the past decade, wind energy has gained more attention in the world. However, owing to its indirectness and volatility properties, wind power penetration has increased the difficulty and complexity in dispatching and planning of electric power systems. Therefore, it is needed to make the high-precision wind power prediction in order to balance the electrical power. For this purpose, in this study, the prediction performance of linear regression, k-nearest neighbor regression and decision tree regression algorithms is compared in detail. k-nearest neighbor regression algorithm provides lower coefficient of determination values, while decision tree regression algorithm produces lower mean absolute error values. In addition, the meteorological parameters of wind speed, wind direction, barometric pressure and air temperature are evaluated in terms of their importance on the wind power parameter. The biggest importance factor is achieved by wind speed parameter. In consequence, many useful assessments are made for wind power predictions.

An emerging optimisation problem from the real-world applications, named the multi-point dynamic aggregation (MPDA) problem, has become one of the active research topics of the multi-robot system. This paper focuses on a multi-objective MPDA problem which is to design an execution plan of the robots to minimise the number of robots and the maximal completion time of all the tasks. The strongly-coupled relationships among robots and tasks, the redundancy of the MPDA encoding, and the variable-size decision space of the MO-MPDA problem posed extra challenges for addressing the problem effectively. To address the above issues, we develop a hybrid decomposition-based multi-objective evolutionary algorithm (HDMOEA) using $ \varepsilon $-constraint method. It selects the maximal completion time of all tasks as the main objective, and converted the other objective into constraints. HDMOEA decomposes a MO-MPDA problem into a series of scalar constrained optimization subproblems by assigning each subproblem with an upper bound robot number. All the subproblems are optimized simultaneously with the transferring knowledge from other subproblems. Besides, we develop a hybrid population initialisation mechanism to enhance the quality of initial solutions, and a reproduction mechanism to transmit effective information and tackle the encoding redundancy. Experimental results show that the proposed HDMOEA method significantly outperforms the state-of-the-art methods in terms of several most-used metrics.

Very recently, a variety of vision transformer architectures for dense prediction tasks have been proposed and they show that the design of spatial attention is critical to their success in these tasks. In this work, we revisit the design of the spatial attention and demonstrate that a carefully-devised yet simple spatial attention mechanism performs favourably against the state-of-the-art schemes. As a result, we propose two vision transformer architectures, namely, Twins-PCPVT and Twins-SVT. Our proposed architectures are highly-efficient and easy to implement, only involving matrix multiplications that are highly optimized in modern deep learning frameworks. More importantly, the proposed architectures achieve excellent performance on a wide range of visual tasks including imagelevel classification as well as dense detection and segmentation. The simplicity and strong performance suggest that our proposed architectures may serve as stronger backbones for many vision tasks. Our code will be released soon at https://github.com/Meituan-AutoML/Twins .