With the rapid development of cloud computing and mobile Internet, there have been a variety of network attacks, among which distributed denial of service (DDoS) is one of the most fatal attacks. Traditional machine l...
详细信息
Since complex data sets and large data magnitude in modern big data require professional analysis tools to achieve analysis, based on this, an unstructured big data analysis algorithm of communicationnetwork based on...
详细信息
In 2020 due to the pandemic of COVID-19, we needed to adapt to the situation and control the amount of people inside buildings to prevent the spread of the virus. Crowd-counting using WiFi is a good approach consideri...
详细信息
ISBN:
(纸本)9798350393156
In 2020 due to the pandemic of COVID-19, we needed to adapt to the situation and control the amount of people inside buildings to prevent the spread of the virus. Crowd-counting using WiFi is a good approach considering the WiFi ubiquity. This paper compares the performance of different machinelearning and Deep learning algorithms for measuring the occupancy level of the room by using WiFi signals, e.g., Naive Bayes, K-Nearest Neighbor (KNN), Linear Discriminant Classifier (LDC), Quadratic Discriminant Classifier (QDC), Support Vector machines (SVM), and 1 Dimension Convolutional Neural network (1DCNN), obtaining the best accuracy of 91.67% using SVM. In addition, we compare the performance by counting the number of people inside the room, with an accuracy of 93.41% applying an SVM strategy.
Local steps are crucial for Federated learning (FL) algorithms and have witnessed great empirical success in reducing communication costs and improving the generalization performance of deep neural networks. However, ...
详细信息
Local steps are crucial for Federated learning (FL) algorithms and have witnessed great empirical success in reducing communication costs and improving the generalization performance of deep neural networks. However, there are limited studies on the effect of local steps on heterogeneous FL. A few works investigate this problem from the optimization perspective. Woodworth et al. (2020a) showed that the iteration complexity of Local SGD, the most popular FL algorithm, is dominated by the baseline mini-batch SGD, which does not show the benefits of local steps. In addition, Levy (2023) proposed a new local update method that provably benefits over mini-batch SGD. However, in the same setting, there is still no work analyzing the effects of local steps to generalization in a heterogeneous FL setting. Motivated by our experimental findings where Local SGD learns more distinguishing features than parallel SGD, this paper studies the generalization benefits of local steps from a feature learning perspective. We propose a novel federated data model that exhibits a new form of data heterogeneity, under which we show that a convolutional neural network (CNN) trained by GD with global updates will miss some pattern-related features, while the network trained by GD with local updates can learn all features in polynomial time. Consequently, local steps help CNN generalize better in our data model. In a different parameter setting, we also prove that Local GD with one-shot model averaging can learn all features and generalize well in all clients. Our experimental results also confirm the benefits of local steps in improving test accuracy on real-world data.
Many deep learning applications benefit from using large models with billions of parameters. Training these models is notoriously expensive due to the need for specialized HPC clusters. In this work, we consider alter...
详细信息
Many deep learning applications benefit from using large models with billions of parameters. Training these models is notoriously expensive due to the need for specialized HPC clusters. In this work, we consider alternative setups for training large models: using cheap "preemptible" instances or pooling existing resources from multiple regions. We analyze the performance of existing model-parallel algorithms in these conditions and find configurations where training larger models becomes less communication-intensive. Based on these findings, we propose SWARM parallelism(1), a model-parallel training algorithm designed for poorly connected, heterogeneous and unreliable devices. SWARM creates temporary randomized pipelines between nodes that are rebalanced in case of failure. We empirically validate our findings and compare SWARM parallelism with existing large-scale training approaches. Finally, we combine our insights with compression strategies to train a large Transformer language model with 1B shared parameters (approximate to 13B before sharing) on preemptible T4 GPUs with less than 200Mb/s network.
We study multitask online learning in a setting where agents can only exchange information with their neighbors on a given arbitrary communicationnetwork. We introduce MT-CO2OL, a decentralized algorithm for this set...
详细信息
We study multitask online learning in a setting where agents can only exchange information with their neighbors on a given arbitrary communicationnetwork. We introduce MT-CO2OL, a decentralized algorithm for this setting whose regret depends on the interplay between the task similarities and the network structure. Our analysis shows that the regret of MT-CO2OL is never worse (up to constants) than the bound obtained when agents do not share information. On the other hand, our bounds significantly improve when neighboring agents operate on similar tasks. In addition, we prove that our algorithm can be made differentially private with a negligible impact on the regret. Finally, we provide experimental support for our theory.
Modern communication relies heavily on mobile cellular networks, which can meet the increasing demands of users for higher data speeds, lower latency, and better communication quality. In order to improve communicatio...
详细信息
With the increasing number of wireless communication devices, the electromagnetic spectrum is becoming increasingly crowded. In order to enhance the ability in monitoring the electromagnetic spectrum and improve the e...
详细信息
Wireless communication is one of the most substantial types of communication nowadays, leading to the IoT revolution and increasing demand for IoT networks. On the other hand, this demand also leads to an increase in ...
详细信息
Human language has many functions. Our communication on social media carries information about how we relate to ourselves and others, that is our identity, and we adjust our language to become more similar to our comm...
详细信息
ISBN:
(纸本)9798400704093
Human language has many functions. Our communication on social media carries information about how we relate to ourselves and others, that is our identity, and we adjust our language to become more similar to our community - in the same way as we dress and style and act to show our commitment to the groups we belong to. Within a community, members adopt the community's language, and the common language becomes a unifying factor. In this paper, we explore the possibilities of identifying linguistic alignment - that individuals adjust their language to become more similar to their conversation partners in a community. We use machinelearning to detect linguistic alignment to a number of different ideologies, communities, and subcultures. We use two different approaches: transfer learning with RoBERTa and traditional machinelearning using Random forest and feature selection.
暂无评论