Distributed Training for Multi-Layer Neural Networks by Consensus

Distributed Optimization of Graph Convolutional Network Using Subgraph Variance

In recent years, distributed graph convolutional networks (GCNs) training frameworks have achieved great success in learning the representation of graph-structured data with large sizes. However, existing distributed GCN training frameworks require enormous communication costs since a multitude of dependent graph data need to be transmitted from other processors. To address this issue, we propose a graph augmentation-based distributed GCN framework (GAD). In particular, GAD has two main components: GAD-Partition and GAD-Optimizer. We first propose an augmentation-based graph partition (GAD-Partition) that can divide the input graph into augmented subgraphs to reduce communication by selecting and storing as few significant vertices of other processors as possible. To further speed up distributed GCN training and improve the quality of the training result, we design a subgraph variance-based importance calculation formula and propose a novel weighted global consensus method, collectively referred to as GAD-Optimizer. This optimizer adaptively adjusts the importance of subgraphs to reduce the effect of extra variance introduced by GAD-Partition on distributed GCN training. Extensive experiments on four large-scale real-world datasets demonstrate that our framework significantly reduces the communication overhead ( ≈ 50% ), improves the convergence speed ( ≈ 2 × ) of distributed GCN training, and obtains a slight gain in accuracy ( ≈ 0.45% ) based on minimal redundancy compared to the state-of-the-art methods.

Consensus-Based Cooperative Algorithms for Training Over Distributed Data Sets Using Stochastic Gradients

In this article, distributed algorithms are proposed for training a group of neural networks with private data sets. Stochastic gradients are utilized in order to eliminate the requirement for true gradients. To obtain a universal model of the distributed neural networks trained using local data sets only, consensus tools are introduced to derive the model toward the optimum. Most of the existing works employ diminishing learning rates, which are often slow and impracticable for online learning, while constant learning rates are studied in some recent works, but the principle for choosing the rates is not well established. In this article, constant learning rates are adopted to empower the proposed algorithms with tracking ability. Under mild conditions, the convergence of the proposed algorithms is established by exploring the error dynamics of the connected agents, which provides an upper bound for selecting the constant learning rates. Performances of the proposed algorithms are analyzed with and without gradient noises, in the sense of mean square error (MSE). It is proved that the MSE converges with bounded errors determined by the gradient noises, and the MSE converges to zero if the gradient noises are absent. Simulation results are provided to validate the effectiveness of the proposed algorithms.

Differential Graphical Games for Constrained Autonomous Vehicles Based on Viability Theory

This article proposes an optimal-distributed control protocol for multivehicle systems with an unknown switching communication graph. The optimal-distributed control problem is formulated to differential graphical games, and the Pareto optimum to multiplayer games is sought based on the viability theory and reinforcement learning techniques. The viability theory characterizes the controllability of a wide range of constrained nonlinear systems; and the viability kernel and the capture basin are the pillars of the viability theory. The capture basin is the set of all initial states, in which there exist control strategies that enable the states to reach the target in finite time while remaining inside a set before reaching the target. In this regard, the feasible learning region is characterized by the reinforcement learner. In addition, the approximation of the capture basin provides the learner with prior knowledge. Unlike the existing works that employ the viability theory to solve control problems with only one agent and differential games with only two players, the viability theory, in this article, is utilized to solve multiagent control problems and multiplayer differential games. The distributed control law is composed of two parts: 1) the approximation of the capture basin and 2) reinforcement learning, which are computed offline and online, respectively. The convergence properties of the parameters' estimation errors in reinforcement learning are proved, and the convergence of the control policy to the Pareto optimum of the differential graphical game is discussed. The guaranteed approximation results of the capture basin are provided and the simulation results of the differential graphical game are provided for multivehicle systems with the proposed distributed control policy.

Deep Learning Adapted to Differential Neural Networks Used as Pattern Classification of Electrophysiological Signals

This manuscript presents the design of a deep differential neural network (DDNN) for pattern classification. First, we proposed a DDNN topology with three layers, whose learning laws are derived from a Lyapunov analysis, justifying local asymptotic convergence of the classification error and the weights of the DDNN. Then, an extension to include an arbitrary number of hidden layers in the DDNN is analyzed. The learning laws for this general form of the DDNN offer a contribution to the deep learning framework for signal classification with biological nature and dynamic structures. The DDNN is used to classify electroencephalographic signals from volunteers that perform an identification graphical test. The classification results show exponential growth in the signal classification accuracy from 82 percent with one layer to 100 percent with three hidden layers. Working with DDNN instead of static deep neural networks (SDNN) represents a set of advantages, such as processing time and training period reduction up to almost 100 times, and the increment of the classification accuracy while working with less hidden layers than working with SDNN, which are highly dependent on their topology and the number of neurons in each layer. The DDNN employed fewer neurons due to the induced feedback characteristic.

Ternary Compression for Communication-Efficient Federated Learning

Learning over massive data stored in different locations is essential in many real-world applications. However, sharing data is full of challenges due to the increasing demands of privacy and security with the growing use of smart mobile devices and Internet of thing (IoT) devices. Federated learning provides a potential solution to privacy-preserving and secure machine learning, by means of jointly training a global model without uploading data distributed on multiple devices to a central server. However, most existing work on federated learning adopts machine learning models with full-precision weights, and almost all these models contain a large number of redundant parameters that do not need to be transmitted to the server, consuming an excessive amount of communication costs. To address this issue, we propose a federated trained ternary quantization (FTTQ) algorithm, which optimizes the quantized networks on the clients through a self-learning quantization factor. Theoretical proofs of the convergence of quantization factors, unbiasedness of FTTQ, as well as a reduced weight divergence are given. On the basis of FTTQ, we propose a ternary federated averaging protocol (T-FedAvg) to reduce the upstream and downstream communication of federated learning systems. Empirical experiments are conducted to train widely used deep learning models on publicly available data sets, and our results demonstrate that the proposed T-FedAvg is effective in reducing communication costs and can even achieve slightly better performance on non-IID data in contrast to the canonical federated learning algorithms.

Distributed Heuristic Adaptive Neural Networks With Variance Reduction in Switching Graphs

This article proposes a distributed adaptive training method for neural networks in switching communication graphs to deal with the problems concerned with massive data or privacy-related data. First, the stochastic variance reduced gradient (SVRG) is used for the training of neural networks. Then, the authors propose a heuristic adaptive consensus algorithm for distributed training, which adaptively adjusts the weighted connectivity matrix based on the performance of each agent over the communication graph. Furthermore, it is proved that the proposed distributed heuristic adaptive neural networks ensure the convergence of all the agents to the optimum with a single communication among connected neighbors after every training step, which is also suitable for switching graphs. This theorem is verified by the simulation, which gives the results that fewer iterations are required for all agents to reach the optimum using the proposed heuristic adaptive consensus algorithm, and the SVRG can greatly decrease the fluctuations caused by the stochastic gradient and improve its performance with only a little extra computational cost.

SaS-BCI: a new strategy to predict image memorability and use mental imagery as a brain-based biometric authentication

Security authentication is one of the most important levels of information security. Nowadays, human biometric techniques are the most secure methods for authentication purposes that cover the problems of older types of authentication like passwords and pins. There are many advantages of recent biometrics in terms of security; however, they still have some disadvantages. Progresses in technology made some specific devices, which make it possible to copy and make a fake human biometric because they are all visible and touchable. According to this matter, there is a need for a new biometric to cover the issues of other types. Brainwave is human data, which uses them as a new type of security authentication that has engaged many researchers. There are some research and experiments, which are investigating and testing EEG signals to find the uniqueness of human brainwave. Some researchers achieved high accuracy rates in this area by applying different signal acquisition techniques, feature extraction and classifications using Brain–Computer Interface (BCI). One of the important parts of any BCI processes is the way that brainwaves could be acquired and recorded. A new Signal Acquisition Strategy is presented in this paper for the process of authorization and authentication of brain signals specifically. This is to predict image memorability from the user’s brain to use mental imagery as a visualization pattern for security authentication. Therefore, users can authenticate themselves with visualizing a specific picture in their minds. In conclusion, we can see that brainwaves can be different according to the mental tasks, which it would make it harder using them for authentication process. There are many signal acquisition strategies and signal processing for brain-based authentication that by using the right methods, a higher level of accuracy rate could be achieved which is suitable for using brain signal as another biometric security authentication.