Differentiable forest is an ensemble of decision trees with full differentiability. Its simple tree structure is easy to use and explain. With full differentiability, it would be trained in the end-to-end learning framework with gradient-based optimization method. In this paper, we propose tree attention block(TAB) in the framework of differentiable forest. TAB block has two operations, squeeze and regulate. The squeeze operation would extract the characteristic of each tree. The regulate operation would learn nonlinear relations between these trees. So TAB block would learn the importance of each tree and adjust its weight to improve accuracy. Our experiment on large tabular dataset shows attention augmented differentiable forest would get comparable accuracy with gradient boosted decision trees(GBDT), which is the state-of-the-art algorithm for tabular datasets. And on some datasets, our model has higher accuracy than best GBDT libs (LightGBM, Catboost, and XGBoost). Differentiable forest model supports batch training and batch size is much smaller than the size of training set. So on larger data sets, its memory usage is much lower than GBDT model. The source codes are available at https://github.com/closest-git/QuantumForest.

Data augmentation is a widely used training trick in deep learning to improve the network generalization ability. Despite many encouraging results, several recent studies did point out limitations of the conventional data augmentation scheme in certain scenarios, calling for a better theoretical understanding of data augmentation. In this work, we develop a comprehensive analysis that reveals pros and cons of data augmentation. The main limitation of data augmentation arises from the data bias, i.e. the augmented data distribution can be quite different from the original one. This data bias leads to a suboptimal performance of existing data augmentation methods. To this end, we develop two novel algorithms, termed "AugDrop" and "MixLoss", to correct the data bias in the data augmentation. Our theoretical analysis shows that both algorithms are guaranteed to improve the effect of data augmentation through the bias correction, which is further validated by our empirical studies. Finally, we propose a generic algorithm "WeMix" by combining AugDrop and MixLoss, whose effectiveness is observed from extensive empirical evaluations.

Federated Learning (FL) is a method of training machine learning models on private data distributed over a large number of possibly heterogeneous clients such as mobile phones and IoT devices. In this work, we propose a new federated learning framework named HeteroFL to address heterogeneous clients equipped with very different computation and communication capabilities. Our solution can enable the training of heterogeneous local models with varying computation complexities and still produce a single global inference model. For the first time, our method challenges the underlying assumption of existing work that local models have to share the same architecture as the global model. We demonstrate several strategies to enhance FL training and conduct extensive empirical evaluations, including five computation complexity levels of three model architecture on three datasets. We show that adaptively distributing subnetworks according to clients' capabilities is both computation and communication efficient. Mobile devices and the Internet of Things (IoT) devices are becoming the primary computing resource for billions of users worldwide (Lim et al., 2020).

While unsupervised domain translation (UDT) has seen a lot of success recently, we argue that allowing its translation to be mediated via categorical semantic features could enable wider applicability. In particular, we argue that categorical semantics are important when translating between domains with multiple object categories possessing distinctive styles, or even between domains that are simply too different but still share high-level semantics. We propose a method to learn, in an unsupervised manner, categorical semantic features (such as object labels) that are invariant of the source and target domains. We show that conditioning the style of a unsupervised domain translation methods on the learned categorical semantics leads to a considerably better high-level features preservation on tasks such as MNIST$\leftrightarrow$SVHN and to a more realistic stylization on Sketches$\to$Reals.

We present a new method for score-based adversarial attack, where the attacker queries the loss-oracle of the target model. Our method employs a parameterized search space with a structure that captures the relationship of the gradient of the loss function. We show that searching over the structured space can be approximated by a time-varying contextual bandits problem, where the attacker takes feature of the associated arm to make modifications of the input, and receives an immediate reward as the reduction of the loss function. The time-varying contextual bandits problem can then be solved by a Bayesian optimization procedure, which can take advantage of the features of the structured action space. The experiments on ImageNet and the Google Cloud Vision API demonstrate that the proposed method achieves the state of the art success rates and query efficiencies for both undefended and defended models. Although deep learning has many applications, it is known that neural networks are vulnerable to adversarial examples, which are small perturbations of inputs that can fool neural networks into making wrong predictions (Szegedy et al., 2014). While adversarial noise can easily be found when the neural models are known (referred to as white-box attack) (Kurakin et al., 2016). However, in real world scenarios models are often unknown, this situation is referred to as black-box attack.

In this paper, we present a novel approach to approximate the gain function of the feedback particle filter (FPF). The exact gain function is the solution of a Poisson equation involving a probability-weighted Laplacian. The numerical problem is to approximate the exact gain function using only finitely many particles sampled from the probability distribution. Inspired by the recent success of the deep learning methods, we represent the gain function as a gradient of the output of a neural network. Thereupon considering a certain variational formulation of the Poisson equation, an optimization problem is posed for learning the weights of the neural network. A stochastic gradient algorithm is described for this purpose. The proposed approach has two significant properties/advantages: (i) The stochastic optimization algorithm allows one to process, in parallel, only a batch of samples (particles) ensuring good scaling properties with the number of particles; (ii) The remarkable representation power of neural networks means that the algorithm is potentially applicable and useful to solve high-dimensional problems. We numerically establish these two properties and provide extensive comparison to the existing approaches.

When using deep neural networks to operate safety-critical systems, assessing the sensitivity of the network outputs when subject to uncertain inputs is of paramount importance. Such assessment is commonly done using reachability analysis or robustness certification. However, certification techniques typically ignore localization information, while reachable set methods can fail to issue robustness guarantees. Furthermore, many advanced methods are either computationally intractable in practice or restricted to very specific models. In this paper, we develop a data-driven optimization-based method capable of simultaneously certifying the safety of network outputs and localizing them. The proposed method provides a unified assessment framework, as it subsumes state-of-the-art reachability analysis and robustness certification. The method applies to deep neural networks of all sizes and structures, and to random input uncertainty with a general distribution. We develop sufficient conditions for the convexity of the underlying optimization, and for the number of data samples to certify and localize the outputs with overwhelming probability. We experimentally demonstrate the efficacy and tractability of the method on a deep ReLU network.

The goal of this work is to shed light on the remarkable phenomenon of transition to linearity of certain neural networks as their width approaches infinity. We show that the transition to linearity of the model and, equivalently, constancy of the (neural) tangent kernel (NTK) result from the scaling properties of the norm of the Hessian matrix of the network as a function of the network width. We present a general framework for understanding the constancy of the tangent kernel via Hessian scaling applicable to the standard classes of neural networks. Our analysis provides a new perspective on the phenomenon of constant tangent kernel, which is different from the widely accepted "lazy training". Furthermore, we show that the transition to linearity is not a general property of wide neural networks and does not hold when the last layer of the network is non-linear. It is also not necessary for successful optimization by gradient descent.

Bootstrapping has been a primary tool for uncertainty quantification, and their theoretical and computational properties have been investigated in the field of statistics and machine learning. However, due to its nature of repetitive computations, the computational burden required to implement bootstrap procedures for the neural network is painfully heavy, and this fact seriously hurdles the practical use of these procedures on the uncertainty estimation of modern deep learning. To overcome the inconvenience, we propose a procedure called Neural Bootstrapper (NeuBoots). We reveal that the NeuBoots stably generate valid bootstrap samples that coincide with the desired target samples with minimal extra computational cost compared to traditional bootstrapping. Consequently, NeuBoots makes it feasible to construct bootstrap confidence intervals of outputs of neural networks and quantify their predictive uncertainty. We also suggest NeuBoots for deep convolutional neural networks to consider its utility in image classification tasks, including calibration, detection of out-of-distribution samples, and active learning. Empirical results demonstrate that NeuBoots is significantly beneficial for the above purposes. Since the introduction of the nonparametric bootstrap (Efron, 1979), bootstrap (or bagging) procedures have been commonly used as a primary tool in quantifying uncertainty lying on statistical inference, e.g.

A crucial problem in neural networks is to select the most appropriate number of hidden neurons and obtain tight statistical risk bounds. In this work, we present a new perspective towards the bias-variance tradeoff in neural networks. As an alternative to selecting the number of neurons, we theoretically show that $L_1$ regularization can control the generalization error and sparsify the input dimension. In particular, with an appropriate $L_1$ regularization on the output layer, the network can produce a statistical risk that is near minimax optimal. Moreover, an appropriate $L_1$ regularization on the input layer leads to a risk bound that does not involve the input data dimension. Our analysis is based on a new amalgamation of dimension-based and norm-based complexity analysis to bound the generalization error. A consequent observation from our results is that an excessively large number of neurons do not necessarily inflate generalization errors under a suitable regularization.