Federated learning enables training a global model from data located at the client nodes, without data sharing and moving client data to a centralized server. Performance of federated learning in a multi-access edge computing (MEC) network suffers from slow convergence due to heterogeneity and stochastic fluctuations in compute power and communication link qualities across clients. We propose a novel coded computing framework, CodedFedL, that injects structured coding redundancy into federated learning for mitigating stragglers and speeding up the training procedure. CodedFedL enables coded computing for non-linear federated learning by efficiently exploiting distributed kernel embedding via random Fourier features that transforms the training task into computationally favourable distributed linear regression. Furthermore, clients generate local parity datasets by coding over their local datasets, while the server combines them to obtain the global parity dataset. Gradient from the global parity dataset compensates for straggling gradients during training, and thereby speeds up convergence. For minimizing the epoch deadline time at the MEC server, we provide a tractable approach for finding the amount of coding redundancy and the number of local data points that a client processes during training, by exploiting the statistical properties of compute as well as communication delays. We also characterize the leakage in data privacy when clients share their local parity datasets with the server. We analyze the convergence rate and iteration complexity of CodedFedL under simplifying assumptions, by treating CodedFedL as a stochastic gradient descent algorithm. Furthermore, we conduct numerical experiments using practical network parameters and benchmark datasets, where CodedFedL speeds up the overall training time by up to $15\times$ in comparison to the benchmark schemes.

Catastrophic forgetting (CF) happens whenever a neural network overwrites past knowledge while being trained on new tasks. Common techniques to handle CF include regularization of the weights (using, e.g., their importance on past tasks), and rehearsal strategies, where the network is constantly re-trained on past data. Generative models have also been applied for the latter, in order to have endless sources of data. In this paper, we propose a novel method that combines the strengths of regularization and generative-based rehearsal approaches. Our generative model consists of a normalizing flow (NF), a probabilistic and invertible neural network, trained on the internal embeddings of the network. By keeping a single NF conditioned on the task, we show that our memory overhead remains constant. In addition, exploiting the invertibility of the NF, we propose a simple approach to regularize the network's embeddings with respect to past tasks. We show that our method performs favorably with respect to state-of-the-art approaches in the literature, with bounded computational power and memory overheads.

This paper proves that the episodic learning environment of every finite-horizon decision task has a unique steady state under any behavior policy, and that the marginal distribution of the agent's input indeed approaches to the steady-state distribution in essentially all episodic learning processes. This observation supports an interestingly reversed mindset against conventional wisdom: While steady states are usually presumed to exist in continual learning and are considered less relevant in episodic learning, it turns out they are guaranteed to exist for the latter. Based on this insight, the paper further develops connections between episodic and continual RL for several important concepts that have been separately treated in the two RL formalisms. Practically, the existence of unique and approachable steady state enables a general, reliable, and efficient way to collect data in episodic RL tasks, which the paper applies to policy gradient algorithms as a demonstration, based on a new steady-state policy gradient theorem. The paper also proposes and empirically evaluates a perturbation method that facilitates rapid mixing in real-world tasks.

We model the behavioral biases of human decision-making in securing interdependent systems and show that such behavioral decision-making leads to a suboptimal pattern of resource allocation compared to non-behavioral (rational) decision-making. We provide empirical evidence for the existence of such behavioral bias model through a controlled subject study with 145 participants. We then propose three learning techniques for enhancing decision-making in multi-round setups. We illustrate the benefits of our decision-making model through multiple interdependent real-world systems and quantify the level of gain compared to the case in which the defenders are behavioral. We also show the benefit of our learning techniques against different attack models. We identify the effects of different system parameters on the degree of suboptimality of security outcomes due to behavioral decision-making.

Anomaly detection is an important research problem because anomalies often contain critical insights for understanding the unusual behavior in data. One type of anomaly detection approach is dependency-based, which identifies anomalies by examining the violations of the normal dependency among variables. These methods can discover subtle and meaningful anomalies with better interpretation. Existing dependency-based methods adopt different implementations and show different strengths and weaknesses. However, the theoretical fundamentals and the general process behind them have not been well studied. This paper proposes a general framework, DepAD, to provide a unified process for dependency-based anomaly detection. DepAD decomposes unsupervised anomaly detection tasks into feature selection and prediction problems. Utilizing off-the-shelf techniques, the DepAD framework can have various instantiations to suit different application domains. Comprehensive experiments have been conducted over one hundred instantiated DepAD methods with 32 real-world datasets to evaluate the performance of representative techniques in DepAD. To show the effectiveness of DepAD, we compare two DepAD methods with nine state-of-the-art anomaly detection methods, and the results show that DepAD methods outperform comparison methods in most cases. Through the DepAD framework, this paper gives guidance and inspiration for future research of dependency-based anomaly detection and provides a benchmark for its evaluation.

The trace of matrix function, far from being only of theoretical interest, appears in many practical applications of linear algebra. To name a few, it has applications in machine learning, computational chemistry, biology, statistics, finance, and many others [1, 6, 13, 14, 24, 26, 31, 49, 52, 53]. While the problem of estimating some spectral quantities dates back to decades, many fast classical algorithms have been developed recently [7, 28, 29, 38, 47, 58, 61], highlighting the importance of spectral sums in many numerical problems. The spectral sum is defined as the sum of the eigenvalues of a matrix after a given function is applied to them. Oftentimes, the matrix will be symmetric positive definite (SPD), but there are cases where this assumption is relaxed. As an example, the logarithm of the determinant is perhaps the most common example of spectral sum, as the determinant is one of the most important properties associated with a matrix. However, the standard definition does not offer an efficient way of computing it. Remarkably, it is often the case that the logarithm of the determinant is the quantity that is effectively needed in the applications, which is much more amenable to estimation.

Different theories have been proposed model's ability to rank a well-organized document to describe the properties that contribute to higher than its noisy counterparts created by discourse coherence and some have been integrated corrupting sentence order in the original document with computational models for empirical (binary discrimination task), and neural evaluation. A popular approach is the entitybased models have achieved remarkable accuracy on model which hypothesizes that coherence this task. Recent efforts have targeted additional can be assessed in terms of the distribution of tasks such as recovering the correct sentence and transitions between entities in a text - by order (Logeswaran et al., 2018; Cui et al., 2018), constructing an entity-grid (Egrid) representation evaluating on realistic data (Lai and Tetreault, (Barzilay and Lapata, 2005, 2008), building 2018; Farag and Yannakoudakis, 2019) and on Centering Theory (Grosz et al., 1995). Subsequent focusing on open-domain models of coherence work has adapted and further extended (Li and Jurafsky, 2017; Xu et al., 2019). Egrid representations (Filippova and Strube, However, less attention has been directed to 2007; Burstein et al., 2010; Elsner and Charniak, investigating and analyzing the properties of coherence 2011; Guinaudeau and Strube, 2013). Other that current models can capture, nor what research has focused on syntactic patterns knowledge is encoded in their representations and that cooccur in text (Louis and Nenkova, how it might relate to aspects of coherence.

Uncertainty quantification (UQ) plays a pivotal role in reduction of uncertainties during both optimization and decision making processes. It can be applied to solve a variety of real-world applications in science and engineering. Bayesian approximation and ensemble learning techniques are two most widely-used UQ methods in the literature. In this regard, researchers have proposed different UQ methods and examined their performance in a variety of applications such as computer vision (e.g., self-driving cars and object detection), image processing (e.g., image restoration), medical image analysis (e.g., medical image classification and segmentation), natural language processing (e.g., text classification, social media texts and recidivism risk-scoring), bioinformatics, etc.This study reviews recent advances in UQ methods used in deep learning. Moreover, we also investigate the application of these methods in reinforcement learning (RL). Then, we outline a few important applications of UQ methods. Finally, we briefly highlight the fundamental research challenges faced by UQ methods and discuss the future research directions in this field.

Preview this course - GET COUPON CODE Learn how to build Machine Learning projects in this TensorFlow Course created by The Click Reader. In this course, you will be learning about Scalar as well as Tensors and how to create them using TensorFlow. You will also be learning how to perform various kinds of Tensor operations for manipulating and changing tensor values. You will be performing a total of three Machine Learning projects while learning through this TensorFlow full course: 1. Linear Regression from Scratch You will be learning how to create a Linear Regression model from scratch using TensorFlow. You will be preparing the data, building the model architecture as well as training the model using a custom-made loss function as well as an optimizer.

Created by Lazy Programmer Inc. English [Auto-generated], Indonesian [Auto-generated], 5 more Created by Lazy Programmer Inc. Like the course I just released on Hidden Markov Models, Recurrent Neural Networks are all about learning sequences - but whereas Markov Models are limited by the Markov assumption, Recurrent Neural Networks are not - and as a result, they are more expressive, and more powerful than anything we've seen on tasks that we haven't made progress on in decades. So what's going to be in this course and how will it build on the previous neural network courses and Hidden Markov Models? In the first section of the course we are going to add the concept of time to our neural networks. I'll introduce you to the Simple Recurrent Unit, also known as the Elman unit.