Model-free deep reinforcement learning algorithms are troubled with poor sample efficiency -- learning reliable policies generally requires a vast amount of interaction with the environment. One hypothesis is that catastrophic interference between various segments within the environment is an issue. In this paper, we perform a large-scale empirical study on the presence of catastrophic interference in the Arcade Learning Environment and find that learning particular game segments frequently degrades performance on previously learned segments. In what we term the Memento observation, we show that an identically parameterized agent spawned from a state where the original agent plateaued, reliably makes further progress. This phenomenon is general -- we find consistent performance boosts across architectures, learning algorithms and environments. Our results indicate that eliminating catastrophic interference can contribute towards improved performance and data efficiency of deep reinforcement learning algorithms.

Online learning is boosting the average UK pay packet by £3,640, or £2 per hour, according to a new study. Britons are turning to the internet to learn new skills which is in turn boosting the UK's economic output, according to the poll of 20,000 of the country's citizens. The Google-backed survey indicates that many working people have used internet-based learning to increase their pay or to get a new job. Work-related online learning that makes use of search engines, video and social media could not only benefit staff but employers' productivity and economic heft. The report concludes that the more you learn the more you earn as people's hunger for knowledge and self-improvement is directly impacting salaries and professional progression'What we found is really encouraging – not just for the businesses and organisations that are benefiting from upskilling employees, but in terms of the economy as a whole,' said Polly Mackenzie, chief executive of Demos, which carried out the survey.

Artificial Intelligence applications are expanding into nearly every area of industry including government services, transportation, healthcare, cybersecurity, autonomous systems, finance and more. Forbes includes artificial intelligence as one of the "Hottest Career Paths of 2020 and Beyond." In order to meet the increasing demand for AI professionals, the University of North Texas, a Tier One research university, is offering the only Master of Science degree in artificial intelligence in Texas and one of only a few programs nationwide. The new degree offers students the choice of three concentrations: machine learning, autonomous systems and biomedical engineering. Students will be able to take classes that allow them to explore specific interests in AI and leave the program with marketable skills.

The tech industry currently holds unprecedented power and influence. Its companies have reached vast market capitalizations, employing hundreds of thousands of workers, and reshaping a range of other industries to accommodate its prerogatives, when it doesn't absorb them outright. Its international political reach is expanding, as companies are intervening in political elections and influencing policy and regulations through millions of dollars spent lobbying. On the academic front, work in tech-related fields such as artificial intelligence and machine learning secures funding with relative ease, its merit and necessity seemingly taken for granted. Directly or indirectly, tech companies continue to invest handsomely in creating an attractive image of the industry, the hackers behind the code, and the technologization of society in general.

The Department of Computer Science at Aalborg University's Technical Faculty of IT and Design is looking to appoint a number of Assistant Professors (tenure-track) for its new group at the university's Copenhagen Campus, commencing September 1, 2020 or soon thereafter. In 2020, the Department of Computer Science will begin building a new research group at Aalborg University's Copenhagen Campus. The group will be responsible for the newly approved bachelor's and master's educations in "Software" and will over time build its own research profile and capacity. The Copenhagen group will build upon, and contribute to the Department's broad range of synergistic activities within research and education in the general area of computer science, including curiosity-driven research and targeted research in collaboration with industrial partners, as well as traditional university education, with a unique problem- and project-based focus, and continued education and knowledge dissemination. As Assistant Professor in Computer Science at the Copenhagen Campus you are expected to deliver teaching of an international standard on the new Software education commencing in September 2020.

In it you will find popular algorithms and architectural solutions that are intuitive to any novice developer. The book covers important topics: from learning neural networks, language processing, feedback neural networks, to mathematical prerequisites and the history of artificial intelligence.

We present TaskSet, a dataset of tasks for use in training and evaluating optimizers. TaskSet is unique in its size and diversity, containing over a thousand tasks ranging from image classification with fully connected or convolutional neural networks, to variational autoencoders, to non-volume preserving flows on a variety of datasets. As an example application of such a dataset we explore meta-learning an ordered list of hyperparameters to try sequentially. By learning this hyperparameter list from data generated using TaskSet we achieve large speedups in sample efficiency over random search. Next we use the diversity of the TaskSet and our method for learning hyperparameter lists to empirically explore the generalization of these lists to new optimization tasks in a variety of settings including ImageNet classification with Resnet50 and LM1B language modeling with transformers. As part of this work we have opensourced code for all tasks, as well as ~29 million training curves for these problems and the corresponding hyperparameters.

In recent years, Deep Neural Networks (DNNs) have gained progressive momentum in many areas of machine learning. The layer-by-layer process of DNNs has inspired the development of many deep models, including deep ensembles. The most notable deep ensemble-based model is Deep Forest, which can achieve highly competitive performance while having much fewer hyper-parameters comparing to DNNs. In spite of its huge success in the batch learning setting, no effort has been made to adapt Deep Forest to the context of evolving data streams. In this work, we introduce the Streaming Deep Forest (SDF) algorithm, a high-performance deep ensemble method specially adapted to stream classification. We also present the Augmented Variable Uncertainty (AVU) active learning strategy to reduce the labeling cost in the streaming context. We compare the proposed methods to state-of-the-art streaming algorithms in a wide range of datasets. The results show that by following the AVU active learning strategy, SDF with only 70\% of labeling budget significantly outperforms other methods trained with all instances.

We investigate several confounding factors in the evaluation of optimization algorithms for deep learning. Primarily, we take a deeper look at how adaptive gradient methods interact with the learning rate schedule, a notoriously difficult-to-tune hyperparameter which has dramatic effects on the convergence and generalization of neural network training. We introduce a "grafting" experiment which decouples an update's magnitude from its direction, finding that many existing beliefs in the literature may have arisen from insufficient isolation of the implicit schedule of step sizes. Alongside this contribution, we present some empirical and theoretical retrospectives on the generalization of adaptive gradient methods, aimed at bringing more clarity to this space.

Machine learning systems must adapt to data distributions that evolve over time, in applications ranging from sensor networks and self-driving car perception modules to brain-machine interfaces. We consider gradual domain adaptation, where the goal is to adapt an initial classifier trained on a source domain given only unlabeled data that shifts gradually in distribution towards a target domain. We prove the first non-vacuous upper bound on the error of self-training with gradual shifts, under settings where directly adapting to the target domain can result in unbounded error. The theoretical analysis leads to algorithmic insights, highlighting that regularization and label sharpening are essential even when we have infinite data, and suggesting that self-training works particularly well for shifts with small Wasserstein-infinity distance. Leveraging the gradual shift structure leads to higher accuracies on a rotating MNIST dataset and a realistic Portraits dataset.