Ensembling deep learning models is a shortcut to promote its implementation in new scenarios, which can avoid tuning neural networks, losses and training algorithms from scratch. However, it is difficult to collect sufficient accurate and diverse models through once training. This paper proposes Auto-Ensemble (AE) to collect checkpoints of deep learning model and ensemble them automatically by adaptive learning rate scheduling algorithm. The advantage of this method is to make the model converge to various local optima by scheduling the learning rate in once training. When the number of lo-cal optimal solutions tends to be saturated, all the collected checkpoints are used for ensemble. Our method is universal, it can be applied to various scenarios. Experiment results on multiple datasets and neural networks demonstrate it is effective and competitive, especially on few-shot learning. Besides, we proposed a method to measure the distance among models. Then we can ensure the accuracy and diversity of collected models.

Active Learning for discriminative models has largely been studied with the focus on individual samples, with less emphasis on how classes are distributed or which classes are hard to deal with. In this work, we show that this is harmful. We propose a method based on the Bayes' rule, that can naturally incorporate class imbalance into the Active Learning framework. We derive that three terms should be considered together when estimating the probability of a classifier making a mistake for a given sample; i) probability of mislabelling a class, ii) likelihood of the data given a predicted class, and iii) the prior probability on the abundance of a predicted class. Implementing these terms requires a generative model and an intractable likelihood estimation. Therefore, we train a Variational Auto Encoder (VAE) for this purpose. To further tie the VAE with the classifier and facilitate VAE training, we use the classifiers' deep feature representations as input to the VAE. By considering all three probabilities, among them especially the data imbalance, we can substantially improve the potential of existing methods under limited data budget. We show that our method can be applied to classification tasks on multiple different datasets -- including one that is a real-world dataset with heavy data imbalance -- significantly outperforming the state of the art.

We propose and analyze zeroth-order algorithms for optimization over Riemannian manifolds, where we observe only potentially noisy evaluations of the objective function. Our approach is based on estimating the Riemannian gradient from the objective function evaluations. We consider three settings for the objective function: (i) deterministic and smooth, (ii) stochastic and smooth, and (iii) composition of smooth and non-smooth parts. For each of the setting, we characterize the oracle complexity of our algorithm to obtain appropriately defined notions of $\epsilon$-stationary points. Notably, our complexities are independent of the ambient dimension of the Euclidean space in which the manifold is embedded in, and only depend on the intrinsic dimension of the manifold. As a proof of concept, we demonstrate the applicability of our method to the problem of black-box attacks to deep neural networks, by providing simulation and real-world image data based experimental results.

Learning effective feature interactions is crucial for click-through rate (CTR) prediction tasks in recommender systems. In most of the existing deep learning models, feature interactions are either manually designed or simply enumerated. However, enumerating all feature interactions brings large memory and computation cost. Even worse, useless interactions may introduce unnecessary noise and complicate the training process. In this work, we propose a two-stage algorithm called Automatic Feature Interaction Selection (AutoFIS). AutoFIS can automatically identify all the important feature interactions for factorization models with just the computational cost equivalent to training the target model to convergence. In the \emph{search stage}, instead of searching over a discrete set of candidate feature interactions, we relax the choices to be continuous by introducing the architecture parameters. By implementing a regularized optimizer over the architecture parameters, the model can automatically identify and remove the redundant feature interactions during the training process of the model. In the \emph{re-train stage}, we keep the architecture parameters serving as an attention unit to further boost the performance. Offline experiments on three large-scale datasets (two public benchmarks, one private) demonstrate that the proposed AutoFIS can significantly improve various FM based models. AutoFIS has been deployed onto the training platform of Huawei App Store recommendation service, where a 10-day online A/B test demonstrated that AutoFIS improved the DeepFM model by 20.3\% and 20.1\% in terms of CTR and CVR respectively.

When designing variational autoencoders (VAEs) or other types of latent space models, the dimensionality of the latent space is typically defined upfront. In this process, it is possible that the number of dimensions is under- or overprovisioned for the application at hand. In case the dimensionality is not predefined, this parameter is usually determined using time- and resource-consuming cross-validation. For these reasons we have developed a technique to shrink the latent space dimensionality of VAEs automatically and on-the-fly during training using Generalized ELBO with Constrained Optimization (GECO) and the $L_0$-Augment-REINFORCE-Merge ($L_0$-ARM) gradient estimator. The GECO optimizer ensures that we are not violating a predefined upper bound on the reconstruction error. This paper presents the algorithmic details of our method along with experimental results on five different datasets. We find that our training procedure is stable and that the latent space can be pruned effectively without violating the GECO constraints.

Assuming hardware is the major constraint for enabling real-time mobile intelligence, the industry has mainly dedicated their efforts to developing specialized hardware accelerators for machine learning and inference. This article challenges the assumption. By drawing on a recent real-time AI optimization framework CoCoPIE, it maintains that with effective compression-compiler co-design, it is possible to enable real-time artificial intelligence on mainstream end devices without special hardware. CoCoPIE is a software framework that holds numerous records on mobile AI: the first framework that supports all main kinds of DNNs, from CNNs to RNNs, transformer, language models, and so on; the fastest DNN pruning and acceleration framework, up to 180X faster compared with current DNN pruning on other frameworks such as TensorFlow-Lite; making many representative AI applications able to run in real-time on off-the-shelf mobile devices that have been previously regarded possible only with special hardware support; making off-the-shelf mobile devices outperform a number of representative ASIC and FPGA solutions in terms of energy efficiency and/or performance.

Not only are Deep Neural Networks (DNNs) black box models, but also we frequently conceptualize them as such. We lack good interpretations of the mechanisms linking inputs to outputs. Therefore, we find it difficult to analyze in human-meaningful terms (1) what the network learned and (2) whether the network learned. We present a hierarchical decomposition of the DNN discrete classification map into logical (AND/OR) combinations of intermediate (True/False) classifiers of the input. Those classifiers that can not be further decomposed, called atoms, are (interpretable) linear classifiers. Taken together, we obtain a logical circuit with linear classifier inputs that computes the same label as the DNN. This circuit does not structurally resemble the network architecture, and it may require many fewer parameters, depending on the configuration of weights. In these cases, we obtain simultaneously an interpretation and generalization bound (for the original DNN), connecting two fronts which have historically been investigated separately. Unlike compression techniques, our representation is. We motivate the utility of this perspective by studying DNNs in simple, controlled settings, where we obtain superior generalization bounds despite using only combinatorial information (e.g. no margin information). We demonstrate how to "open the black box" on the MNIST dataset. We show that the learned, internal, logical computations correspond to semantically meaningful (unlabeled) categories that allow DNN descriptions in plain English. We improve the generalization of an already trained network by interpreting, diagnosing, and replacing components the logical circuit that is the DNN.

COVID-19, the disease caused by the SARS-CoV-2 virus, has been declared a pandemic by the World Health Organization, with over 294,000 cases as of March 22, 2020 (WHO, 2020). In this review, we present an overview of recent studies using Machine Learning and, more broadly, Artificial Intelligence, to tackle many aspects of the COVID-19 crisis at different scales including molecular, medical and epidemiological applications. We finish with a discussion of promising future directions of research and the tools and resources needed to facilitate AI research. Executive Summary - There is a broad range of potential applications of AI covering medical and societal challenges created by the COVID-19 pandemic; however, few of them are currently mature enough to show operational impact.

In 2016, Seattle-based startup Turi Inc. was helping almost 100 customers create and manage software that uses machine learning, a powerful type of artificial intelligence. Its technology was so promising that Apple Inc. snapped it up for $200 million. The deal was a triumph for investors and founders, but one backer thought Turi -- and the broader tech industry -- might be better off if the startup had spurned Apple's advances. Matt McIlwain, managing director at Madrona Venture Group, said it's important that at least some emerging tech businesses remain independent, rather than falling into the arms of Apple, Amazon.com "It is economically beneficial to society to have more stand-alone, independent companies. We generally think that's better than just having these companies consolidated into larger ones," McIlwain said.

In 2016, Seattle-based startup Turi Inc. was helping almost 100 customers create and manage software that uses machine learning, a powerful type of artificial intelligence. Its technology was so promising that Apple Inc. snapped it up for $200 million. The deal was a triumph for investors and founders, but one backer thought Turi -- and the broader tech industry -- might be better off if the startup had spurned Apple's advances. Matt McIlwain, managing director at Madrona Venture Group, said it's important that at least some emerging tech businesses remain independent, rather than falling into the arms of Apple, Amazon.com "It is economically beneficial to society to have more stand-alone, independent companies. We generally think that's better than just having these companies consolidated into larger ones," McIlwain said.