Distributed learning techniques such as federated learning have enabled multiple workers to train machine learning models together to reduce the overall training time. However, current distributed training algorithms (centralized or decentralized) suffer from the communication bottleneck on multiple low-bandwidth workers (also on the server under the centralized architecture). Although decentralized algorithms generally have lower communication complexity than the centralized counterpart, they still suffer from the communication bottleneck for workers with low network bandwidth. To deal with the communication problem while being able to preserve the convergence performance, we introduce a novel decentralized training algorithm with the following key features: 1) It does not require a parameter server to maintain the model during training, which avoids the communication pressure on any single peer. 2) Each worker only needs to communicate with a single peer at each communication round with a highly compressed model, which can significantly reduce the communication traffic on the worker. We theoretically prove that our sparsification algorithm still preserves convergence properties. 3) Each worker dynamically selects its peer at different communication rounds to better utilize the bandwidth resources. We conduct experiments with convolutional neural networks on 32 workers to verify the effectiveness of our proposed algorithm compared to seven existing methods. Experimental results show that our algorithm significantly reduces the communication traffic and generally select relatively high bandwidth peers.

Vehicle tracking has become one of the key applications of wireless sensor networks (WSNs) in the fields of rescue, surveillance, traffic monitoring, etc. However, the increased tracking accuracy requires more energy consumption. In this letter, a decentralized vehicle tracking strategy is conceived for improving both tracking accuracy and energy saving, which is based on adjusting the intersection area between the fixed sensing area and the dynamic activation area. Then, two deep reinforcement learning (DRL) aided solutions are proposed relying on the dynamic selection of the activation area radius. Finally, simulation results show the superiority of our DRL aided design.

This paper presents a multi-staged approach to nonmyopic adaptive Gaussian process optimization (GPO) for Bayesian optimization (BO) of unknown, highly complex objective functions that, in contrast to existing nonmyopic adaptive BO algorithms, exploits the notion of macro-actions for scaling up to a further lookahead to match up to a larger available budget. To achieve this, we generalize GP upper confidence bound to a new acquisition function defined w.r.t. a nonmyopic adaptive macro-action policy, which is intractable to be optimized exactly due to an uncountable set of candidate outputs. The contribution of our work here is thus to derive a nonmyopic adaptive epsilon-Bayes-optimal macro-action GPO (epsilon-Macro-GPO) policy. To perform nonmyopic adaptive BO in real time, we then propose an asymptotically optimal anytime variant of our epsilon-Macro-GPO policy with a performance guarantee. We empirically evaluate the performance of our epsilon-Macro-GPO policy and its anytime variant in BO with synthetic and real-world datasets.

Graph based clustering plays an important role in clustering area. Recent studies about graph convolution neural networks have achieved impressive success on graph type data. However, in traditional clustering tasks, the graph structure of data does not exist such that the strategy to construct graph is crucial for performance. In addition, the existing graph auto-encoder based approaches perform poorly on weighted graph, which is widely used in graph based clustering. In this paper, we propose a graph auto-encoder with local structure preserving for general data clustering, which can update the constructed graph adaptively. The adaptive process is designed to utilize the non-Euclidean structure sufficiently. By combining generative model for graph embedding and graph based clustering, a graph auto-encoder with a novel decoder is developed and it performs well in weighted graph used scenarios. Extensive experiments prove the superiority of our model.

Analysing and computing with Gaussian processes arising from infinitely wide neural networks has recently seen a resurgence in popularity. Despite this, many explicit covariance functions of networks with activation functions used in modern networks remain unknown. Furthermore, while the kernels of deep networks can be computed iteratively, theoretical understanding of deep kernels is lacking, particularly with respect to fixed-point dynamics. Firstly, we derive the covariance functions of MLPs with exponential linear units and Gaussian error linear units and evaluate the performance of the limiting Gaussian processes on some benchmarks. Secondly, and more generally, we introduce a framework for analysing the fixed-point dynamics of iterated kernels corresponding to a broad range of activation functions. We find that unlike some previously studied neural network kernels, these new kernels exhibit non-trivial fixed-point dynamics which are mirrored in finite-width neural networks.

Widely popular transformer-based NLP models such as BERT and GPT have enormous capacity trending to billions of parameters. Current execution methods demand brute-force resources such as HBM devices and high speed interconnectivity for data parallelism. In this paper, we introduce a new relay-style execution technique called L2L (layer-to-layer) where at any given moment, the device memory is primarily populated only with the executing layer(s)'s footprint. The model resides in the DRAM memory attached to either a CPU or an FPGA as an entity we call eager param-server (EPS). Unlike a traditional param-server, EPS transmits the model piecemeal to the devices thereby allowing it to perform other tasks in the background such as reduction and distributed optimization. To overcome the bandwidth issues of shuttling parameters to and from EPS, the model is executed a layer at a time across many micro-batches instead of the conventional method of minibatches over whole model. In this paper, we explore a conservative version of L2L that is implemented on a modest Azure instance for BERT-Large running it with a batch size of 32 on a single V100 GPU using less than 8GB memory. Our results show a more stable learning curve, faster convergence, better accuracy and 35% reduction in memory compared to the state-of-the-art baseline. Our method reproduces BERT results on any mid-level GPU that was hitherto not feasible. L2L scales to arbitrary depth without impacting memory or devices allowing researchers to develop affordable devices. It also enables dynamic approaches such as neural architecture search. This work has been performed on GPUs first but also targeted towards high TFLOPS/Watt accelerators such as Graphcore IPUs. The code will soon be available on github.

Medical imaging systems are commonly assessed by use of objective image quality measures. Supervised deep learning methods have been investigated to implement numerical observers for task-based image quality assessment. However, labeling large amounts of experimental data to train deep neural networks is tedious, expensive, and prone to subjective errors. Computer-simulated image data can potentially be employed to circumvent these issues; however, it is often difficult to computationally model complicated anatomical structures, noise sources, and the response of real world imaging systems. Hence, simulated image data will generally possess physical and statistical differences from the experimental image data they seek to emulate. Within the context of machine learning, these differences between the sets of two images is referred to as domain shift. In this study, we propose and investigate the use of an adversarial domain adaptation method to mitigate the deleterious effects of domain shift between simulated and experimental image data for deep learning-based numerical observers (DL-NOs) that are trained on simulated images but applied to experimental ones. In the proposed method, a DL-NO will initially be trained on computer-simulated image data and subsequently adapted for use with experimental image data, without the need for any labeled experimental images. As a proof of concept, a binary signal detection task is considered. The success of this strategy as a function of the degree of domain shift present between the simulated and experimental image data is investigated.

Optimal transport (OT) is a powerful geometric and probabilistic tool for finding correspondences and measuring similarity between two distributions. Yet, its original formulation relies on the existence of a cost function between the samples of the two distributions, which makes it impractical for comparing data distributions supported on different topological spaces. To circumvent this limitation, we propose a novel OT problem, named COOT for CO-Optimal Transport, that aims to simultaneously optimize two transport maps between both samples and features. This is different from other approaches that either discard the individual features by focussing on pairwise distances (e.g. Gromov-Wasserstein) or need to model explicitly the relations between the features. COOT leads to interpretable correspondences between both samples and feature representations and holds metric properties. We provide a thorough theoretical analysis of our framework and establish rich connections with the Gromov-Wasserstein distance. We demonstrate its versatility with two machine learning applications in heterogeneous domain adaptation and co-clustering/data summarization, where COOT leads to performance improvements over the competing state-of-the-art methods.

We propose a unifying view of two different Bayesian inference algorithms, Stochastic Gradient Markov Chain Monte Carlo (SG-MCMC) and Stein Variational Gradient Descent (SVGD), leading to improved and efficient novel sampling schemes. We show that SVGD combined with a noise term can be framed as a multiple chain SG-MCMC method. Instead of treating each parallel chain independently from others, our proposed algorithm implements a repulsive force between particles, avoiding collapse and facilitating a better exploration of the parameter space. We also show how the addition of this noise term is necessary to obtain a valid SG-MCMC sampler, a significant difference with SVGD. Experiments with both synthetic distributions and real datasets illustrate the benefits of the proposed scheme.

The aim of this work is to provide a family of qualitative theories for spatial change in general, and for motion of spatial scenes in particular. To achieve this, we consider a spatio-temporalisation MTALC(Dx), of the well-known ALC(D) family of Description Logics (DLs) with a concrete domain: the MTALC(Dx) concepts are interpreted over infinite k-ary Sigma-trees, with the nodes standing for time points, and Sigma including, additionally to its uses in classical k-ary Sigma-trees, the description of the snapshot of an n-object spatial scene of interest; the roles split into m+n immediate-successor (accessibility) relations, which are serial, irreflexive and antisymmetric, and of which m are general, not necessarily functional, the other n functional; the concrete domain Dx is generated by an RCC8-like spatial Relation Algebra (RA) x, and is used to guide the change by imposing spatial constraints on objects of the "followed" spatial scene, eventually at different time points of the input trees. In order to capture the expressiveness of most modal temporal logics encountered in the literature, we introduce weakly cyclic Terminological Boxes (TBoxes) of MTALC(Dx), whose axioms capture the decreasing property of modal temporal operators. We show the important result that satisfiability of an MTALC(Dx) concept with respect to a weakly cyclic TBox can be reduced to the emptiness problem of a Buchi weak alternating automaton augmented with spatial constraints. In another work, complementary to this one, also submitted to this conference, we thoroughly investigate Buchi automata augmented with spatial constraints, and provide, in particular, a translation of an alternating into a nondeterministic, and an effective decision procedure for the emptiness problem of the latter.