In recent years, hyper-complex deep networks (e.g., quaternion-based) have received increasing interest with applications ranging from image reconstruction to 3D audio processing. Similarly to their real-valued counterparts, quaternion neural networks might require custom regularization strategies to avoid overfitting. In addition, for many real-world applications and embedded implementations there is the need of designing sufficiently compact networks, with as few weights and units as possible. However, the problem of how to regularize and/or sparsify quaternion-valued networks has not been properly addressed in the literature as of now. In this paper we show how to address both problems by designing targeted regularization strategies, able to minimize the number of connections and neurons of the network during training. To this end, we investigate two extensions of $\ell_1$ and structured regularization to the quaternion domain. In our experimental evaluation, we show that these tailored strategies significantly outperform classical (real-valued) regularization strategies, resulting in small networks especially suitable for low-power and real-time applications.

How to best explore in domains with sparse, delayed, and deceptive rewards is an important open problem for reinforcement learning (RL). This paper considers one such domain, the recently-proposed multi-agent benchmark of Pommerman. This domain is very challenging for RL --- past work has shown that model-free RL algorithms fail to achieve significant learning without artificially reducing the environment's complexity. In this paper, we illuminate reasons behind this failure by providing a thorough analysis on the hardness of random exploration in Pommerman. While model-free random exploration is typically futile, we develop a model-based automatic reasoning module that can be used for safer exploration by pruning actions that will surely lead the agent to death. We empirically demonstrate that this module can significantly improve learning.

A key challenge in reinforcement learning (RL) is environment generalization: a policy trained to solve a task in one environment often fails to solve the same task in a slightly different test environment. A common approach to improve inter-environment transfer is to learn policies that are invariant to the distribution of testing environments. However, we argue that instead of being invariant, the policy should identify the specific nuances of an environment and exploit them to achieve better performance. In this work, we propose the 'Environment-Probing' Interaction (EPI) policy, a policy that probes a new environment to extract an implicit understanding of that environment's behavior. Once this environment-specific information is obtained, it is used as an additional input to a task-specific policy that can now perform environment-conditioned actions to solve a task. To learn these EPI-policies, we present a reward function based on transition predictability. Specifically, a higher reward is given if the trajectory generated by the EPI-policy can be used to better predict transitions. We experimentally show that EPI-conditioned task-specific policies significantly outperform commonly used policy generalization methods on novel testing environments.

Deep learning has been shown to achieve impressive results in several domains like computer vision and natural language processing. Deep architectures are typically trained following a supervised scheme and, therefore, they rely on the availability of a large amount of labeled training data to effectively learn their parameters. Neuro-symbolic approaches have recently gained popularity to inject prior knowledge into a deep learner without requiring it to induce this knowledge from data. These approaches can potentially learn competitive solutions with a significant reduction of the amount of supervised data. A large class of neuro-symbolic approaches is based on First-Order Logic to represent prior knowledge, that is relaxed to a differentiable form using fuzzy logic. This paper shows that the loss function expressing these neuro-symbolic learning tasks can be unambiguously determined given the selection of a t-norm generator. When restricted to simple supervised learning, the presented theoretical apparatus provides a clean justification to the popular cross-entropy loss, that has been shown to provide faster convergence and to reduce the vanishing gradient problem in very deep structures. One advantage of the proposed learning formulation is that it can be extended to all the knowledge that can be represented by a neuro-symbolic method, and it allows the development of a novel class of loss functions, that the experimental results show to lead to faster convergence rates than other approaches previously proposed in the literature.

Performing machine learning (ML) and analyzing geospatial data are both hard problems requiring a lot of domain expertise. These limitations have historically meant that one needs to be an expert in both to perform even the most basic analyses, making advances in AI for overhead imagery difficult to achieve. We at CosmiQ Works have asked ourselves: is there anything we can do to reduce this barrier to entry, making it easier to apply machine learning methods to overhead imagery data? Enter Solaris, a new Python library for ML analysis of geospatial data from CosmiQ Works. Solaris builds upon SpaceNet's previous tool suite, SpaceNetUtilities, along with several other CosmiQ projects like BASISS to provide an end-to-end pipeline for geospatial AI. Would you prefer a basic command line interface so you can run a pre-trained model without learning Python?

ODSC is the best community data science event on the planet. There are other events that cover special topics, or industries, etc., but ODSC is comprehensive and totally community-focused: it's the conference to engage, to build, to develop, and to learn from the whole data science community. ODSC West 2019 is one of the largest applied data science conferences in the world. Our speakers include some of the core contributors to many open source tools, libraries, and languages. Attend ODSC West 2019 and learn the latest AI & data science topics, tools, and languages from some of the best and brightest minds in the field.

Next, we examined how segmentation performance varies as a function of underlying disease pathology. The CNN had significantly higher mean Dice scores than at least 1 of the other methods in 16/19 diagnoses and significantly higher than those in both other methods in 5/19 diagnoses (paired 2-tailed t tests, P .05;

AI models capable of reliably guiding driverless cars typically require endless testing and fine-tuning, not to mention computational power out the wazoo. In an effort to bolster AI algorithm training effectiveness and efficiency, Google parent company Alphabet's Waymo is collaborating with DeepMind on techniques inspired by evolutionary biology, the two companies revealed in a blog post this morning. As Waymo explains, AI algorithms self-improve through trial and error. A model is presented with a task that it learns to perform by continually attempting it and adjusting based on the feedback it receives. Performance is heavily dependent on the training regimen -- known as a hyperparemeter schedule -- and finding the best regimen is commonly left to experienced researchers and engineers. They handpick AI models undergoing training, culling the weakest performers and freeing resources to train new algorithms from scratch.

Edge computing brings computation and data storage closer to where it is needed, to improve response times and save bandwidth. Now more AI is being incorporated into edge devices, from IoT devices to smartphones to automobiles, as edge compute power increases and AI algorithms improve. As the edge computer processes, only a subset of data generated by sensors is sent to the cloud, saving on bandwidth and cloud storage costs, according to a recent account in Forbes. Decisions can be made faster using AI models on the edge that have been trained via machine learning in the cloud. A deep learning model deployed at the edge may see slower inferencing.

How can we trust the networks?" They answered: "We know that a network is quite reliable when its inputs come from its training set. But these critical systems will have inputs corresponding to new, often unanticipated situations. There are numerous examples where a network gives poor responses for untrained inputs." David Lorge Parnas followed up on this discussion in his Letter to the Editor (Feb. We wish to point out that machine learning-based systems, including commercial ones performing safety critical tasks, can fail not only under "unanticipated situations" (noted by Lewis and Denning) or "when it encounters data radically different from its training set" (noted by Parnas), but also under normal situations, even on data that is extremely similar to its training set. The Apollo self-driving team confirmed "it might happen" because the system was "deep learning trained." Now, after a further investigation, we have found that in 24 of these 27 failed tests, the 10 random points ...