Being able to explain a prediction as well as having a model that performs well are paramount in many machine learning applications. Deep neural networks have gained momentum recently on the basis of their accuracy, however these are often criticised to be black-boxes. Many authors have focused on proposing methods to explain their predictions. Among these explainability methods, feature attribution methods have been favoured for their strong theoretical foundation: the Shapley value. A limitation of Shapley value is the need to define a baseline (aka reference point) representing the missingness of a feature. In this paper, we present a method to choose a baseline based on a neutrality value: a parameter defined by decision makers at which their choices are determined by the returned value of the model being either below or above it. Based on this concept, we theoretically justify these neutral baselines and find a way to identify them for MLPs. Then, we experimentally demonstrate that for a binary classification task, using a synthetic dataset and a dataset coming from the financial domain, the proposed baselines outperform, in terms of local explanability power, standard ways of choosing them.

With the introduction of SNIP [arXiv:1810.02340v2], it has been demonstrated that modern neural networks can effectively be pruned before training. Yet, its sensitivity criterion has since been criticized for not propagating training signal properly or even disconnecting layers. As a remedy, GraSP [arXiv:2002.07376v1] was introduced, compromising on simplicity. However, in this work we show that by applying the sensitivity criterion iteratively in smaller steps - still before training - we can improve its performance without difficult implementation. As such, we introduce 'SNIP-it'. We then demonstrate how it can be applied for both structured and unstructured pruning, before and/or during training, therewith achieving state-of-the-art sparsity-performance trade-offs. That is, while already providing the computational benefits of pruning in the training process from the start. Furthermore, we evaluate our methods on robustness to overfitting, disconnection and adversarial attacks as well.

Rapid developments in data collecting devices and computation platforms produce an emerging number of learners and data modalities in many scientific domains. We consider the setting in which each learner holds a pair of parametric statistical model and a specific data source, with the goal of integrating information across a set of learners to enhance the prediction accuracy of a specific learner. One natural way to integrate information is to build a joint model across a set of learners that shares common parameters of interest. However, the parameter sharing patterns across a set of learners are not known a priori. Misspecifying the parameter sharing patterns and the parametric statistical model for each learner yields a biased estimator and degrades the prediction accuracy of the joint model. In this paper, we propose a novel framework for integrating information across a set of learners that is robust against model misspecification and misspecified parameter sharing patterns. The main crux is to sequentially incorporates additional learners that can enhance the prediction accuracy of an existing joint model based on a user-specified parameter sharing patterns across a set of learners, starting from a model with one learner. Theoretically, we show that the proposed method can data-adaptively select the correct parameter sharing patterns based on a user-specified parameter sharing patterns, and thus enhances the prediction accuracy of a learner. Extensive numerical studies are performed to evaluate the performance of the proposed method.

Digital twin technology has a huge potential for widespread applications in different industrial sectors such as infrastructure, aerospace, and automotive. However, practical adoptions of this technology have been slower, mainly due to a lack of application-specific details. Here we focus on a digital twin framework for linear single-degree-of-freedom structural dynamic systems evolving in two different operational time scales in addition to its intrinsic dynamic time-scale. Our approach strategically separates into two components -- (a) a physics-based nominal model for data processing and response predictions, and (b) a data-driven machine learning model for the time-evolution of the system parameters. The physics-based nominal model is system-specific and selected based on the problem under consideration. On the other hand, the data-driven machine learning model is generic. For tracking the multi-scale evolution of the system parameters, we propose to exploit a mixture of experts as the data-driven model. Within the mixture of experts model, Gaussian Process (GP) is used as the expert model. The primary idea is to let each expert track the evolution of the system parameters at a single time-scale. For learning the hyperparameters of the `mixture of experts using GP', an efficient framework the exploits expectation-maximization and sequential Monte Carlo sampler is used. Performance of the digital twin is illustrated on a multi-timescale dynamical system with stiffness and/or mass variations. The digital twin is found to be robust and yields reasonably accurate results. One exciting feature of the proposed digital twin is its capability to provide reasonable predictions at future time-steps. Aspects related to the data quality and data quantity are also investigated.

Actor-critic (AC) methods have exhibited great empirical success compared with other reinforcement learning algorithms, where the actor uses the policy gradient to improve the learning policy and the critic uses temporal difference learning to estimate the policy gradient. Under the two time-scale learning rate schedule, the asymptotic convergence of AC has been well studied in the literature. However, the non-asymptotic convergence and finite sample complexity of actor-critic methods are largely open. In this work, we provide a non-asymptotic analysis for two time-scale actor-critic methods under non-i.i.d. setting. We prove that the actor-critic method is guaranteed to find a first-order stationary point (i.e., $\|\nabla J(\boldsymbol{\theta})\|_2^2 \le \epsilon$) of the non-concave performance function $J(\boldsymbol{\theta})$, with $\mathcal{\tilde{O}}(\epsilon^{-2.5})$ sample complexity. To the best of our knowledge, this is the first work providing finite-time analysis and sample complexity bound for two time-scale actor-critic methods.

The UK League Championship (UKLC) League of Legends competition is coming to the BBC, with some of the best esports players in the country fighting it out for a share of a £20,000 prize. Eight teams will compete in a round-robin tournament over the next seven weeks to try and make it to the end-of-season playoffs, and you can watch each match of the competition on BBC iPlayer and the BBC Sport website from Sunday 14 June until Monday 27 July. With so much action in the pipeline, you will need to get up to speed on everything UKLC before the competition kicks off... League of Legends is one of the most popular video games in the world, with millions of monthly players. Two teams of five compete on a map that has three lanes, to try and destroy the enemy's base. In order to do that they will have to kill minions controlled by artificial intelligence and enemy players to increase their power, then destroy the towers outside the base before they can finally make their way to the Nexus, the building that needs to be demolished to win.

"Can you recommend a free self-paced learning path for #machinelearning and #deeplearning?" This is based on my work / teaching students primarily at Oxford University, but I have chosen only free resources here i.e. publicly available. Usual disclaimers apply i.e. the views are my own So, my suggestion is: Use this learning pathway as a guide but shorten it as you want. Try to go on a series of small journeys – each of which you will complete. But overall, try and maintain the sequence and these resources (trust me between them – I don't think you will miss anything!)

Artificial intelligence already helps run government, with cognitive applications doing everything from reducing backlogs and cutting costs to handling tasks we can't easily do on our own, such as predicting fraudulent transactions and identifying criminal suspects via facial recognition. Indeed, while we expect AI-based technology in the years ahead to fundamentally transform how public-sector employees get work done--eliminating some jobs, redesigning countless others, and even creating entirely new professions1--it's already changing the nature of many jobs and revolutionizing facets of government operations. Agencies today face new choices about whether some work should be fully automated, divided among people and machines, or performed by people but enhanced by machines. Our latest report, AI-augmented government, conservatively estimates that simply automating tasks that computers already routinely do could free up 96.7 million federal government working hours annually, potentially saving $3.3 billion. At the high end, we estimate that AI technology could free up as many as 1.2 billion working hours every year, saving $41.1 billion.

This week IBM, Microsoft and Amazon announced that they would suspend the sale of their facial recognition technology to law enforcement agencies. But the moves from the tech giants also illustrate the inherent risks of AI, especially when it comes to bias and the potential for invasion of privacy. Note that there are already indications that Congress will take action to regulate the technology. In the meantime, many cities have already instituted bans, such San Francisco. Because of the advances of deep learning and faster systems for processing enormous amounts of data, facial recognition has certainly seen major strides over the past decade.

Artificial Intelligence And Machine Learning Are More Important Than You Might Think For Mobile ... [ ] Devices AI is a common buzz word these days, but most consumers probably aren't aware how it's interwoven in their everyday lives. Some of us in the analyst and tech press communities may also scoff at how often the term is used for some technologies that hardly resemble true artificial intelligence. That said, there are a few platforms, beyond just powerful data centers, that are a natural for AI processing and the NNs (Neural Networks) that drive them. One of those is AI inferencing (using the AI to infer information, versus training an NN) at the edge, and in your pocket, with a smartphone. As you might imagine, smartphone platforms from Android to Apple vary greatly, but there are common workloads like speech-to-text translation, and recommender engines (like Google Assistant and Siri), that make heavy use of common AI NN models, and they do so on-device for speed and latency advantages.