Search

Multi-Task Learning (MTL) is a learning paradigm in machine learning and its aim is to leverage useful information contained in multiple related tasks to help improve the generalization performance of all the tasks. In this paper, we give a survey for MTL from the perspective of algorithmic modeling, applications and theoretical analyses. For algorithmic modeling, we give a definition of MTL and then classify different MTL algorithms into five categories, including feature learning approach, low-rank approach, task clustering approach, task relation learning approach and decomposition approach as well as discussing the characteristics of each approach. In order to improve the performance of learning tasks further, MTL can be combined with other learning paradigms including semi-supervised learning, active learning, unsupervised learning, reinforcement learning, multi-view learning and graphical models. When the number of tasks is large or the data dimensionality is high, we review online, parallel and distributed MTL models as well as dimensionality reduction and feature hashing to reveal their computational and storage advantages. Many real-world applications use MTL to boost their performance and we review representative works in this paper. Finally, we present theoretical analyses and discuss several future directions for MTL. [ABSTRACT FROM AUTHOR]

The accuracy of artificial intelligence (AI) models is crucial for connected and autonomous vehicles (CAVs). However, in reality, model training under less frequent weather faces the problem of insufficient sampling. Also, in the real world, weather, sunlight, etc., can only change with the speed of the real-time clock, so the traditional sampling process is very slow. Moreover, currently, collective learning, which can make up the limited experience and computing power of a single vehicle, is always introduced to cases where the data from participants have the same structure, wasting massive heterogeneous data from vehicles of different brands. Therefore, in this paper, we propose a quantum collective learning and many-to-many matching game-based scheme in the metaverse for CAVs. The environment is simulated in the metaverse, which has its own time clock system, thereby expanding sample size and speeding up the sampling process. And we quantify the quality of intelligence in collective learning from the perspective of feature diversity. It is the cornerstone of collective learning between heterogeneous vehicles, facilitating maximum utilization of data with different structures. Then, we formulate the distributed vehicles selection problem as a many-to-many matching game and use Gale–Shapely algorithm to solve it. Also, we formulate the spectrum resource allocation problem as a discrete Markov decision process (MDP) and adopt a quantum-inspired reinforcement learning (QRL) algorithm to find the optimal policy to achieve the high revenue of the system. In simulations, the performance of the proposed scheme is compared with existing methods. [ABSTRACT FROM AUTHOR]

With the integration of Artificial Intelligence (AI) and Internet of Things (IoT), the Federated Edge Learning (FEL), a promising computing framework is developing. However, there are still unsolved issues on communication efficiency and data security due to the huge models and unreliable transmission links. To address these issues, this paper proposes a novel federated edge learning system, called LightFed, where the edge nodes upload only vital partial local models, and successfully achieve lightweight communication and model aggregation. First, a novel model aggregation method Model Splitting and Splicing (MSS) and a Selective Parameter Transmission (SPT) scheme are proposed. By detecting the updating gradients of local parameters and filtering significant parameters, selective rotated transmission and efficient aggregation of local models are achieved. Second, a Training Filling Model (TFM) is proposed to infer the total data distribution of edge nodes, and train a filling model to mitigate the unbalanced training data without violating the data privacy of individual users. Moreover, a blockchain-powered confusion transmission mechanism is proposed for defending the attacks from external adversaries and protecting the model information. Finally, extensive experimental results demonstrate that our LightFed significantly outperforms the existing FEL systems in terms of communication efficiency and privacy security. [ABSTRACT FROM AUTHOR]

Artificial intelligence (AI)-powered autonomous vehicles (AVs) can integrate different machine learning (ML) techniques to build up a complex autonomous driving system. However, single AV intelligence is not enough to cope with ever-changing driving environments. The underlying reason is that, with current neural network design and training algorithms, it is challenging for the driving model to generalize to diverse driving environments all at once due to sample inefficiency and the curse of dimensionality. Powerful computing resources and massive amount of data can be used to train a good driving model offline. However, the driving model obtained offline might fail in corner case scenarios. In this paper, we propose an intelligence networking framework among AVs assisted by multi-access edge computing (MEC) with end-to-end learning for demonstration. In this framework, driving road is divided into segments and data is collected for each road segment separately. Assisted by MEC networks, a continuously updated driving model is produced in near real-time for each road segment when the environment changes. By dividing the road into segments, we aim to reduce the burden of generalization since a single model only needs to adapt to a specific road segment. Simulation results show that our solutions can produce updated driving model for each road segment to adapt to environmental changes better than the existing scheme. Upcoming AVs can then adapt to changing environments by downloading the updated driving model. [ABSTRACT FROM AUTHOR]

With 5G already deployed, challenges related to handover exacerbate due to the dense base station deployment operating on a motley of frequencies. In this paper, we present and evaluate a novel data-driven solution, to reduce inter-frequency handover failures (HOFs), hereafter referred to as TORIS (Transmit Power Tuning-based Handover Success Rate Improvement Scheme). TORIS is designed by developing and integrating two sub-solutions. First sub-solution consists of an Artificial Intelligence (AI)-based model to predict inter-frequency HOFs. In this model, we achieve higher than the state-of-the-art accuracy by leveraging two approaches. First, we devise a novel feature set by exploiting domain knowledge gathered from extensive drive test data analysis. Second, we exploit an extensive set of data augmentation techniques to address the class imbalance in training the HOF prediction model. The data augmentation techniques include Chow-Liu Bayesian Network and Generative Adversarial Network further improved by focusing the sampling only on the borderline. We also compare the performance of state-of-the-art AI models for predicting HOFs with and without augmented data. Results show that AdaBoost yields best performance for predicting HOFs. The second sub-solution is a heuristic scheme to tune the transmit (Tx) power of serving and target cells. Unlike the state-of-the-art approaches for HOF reduction that tune cell individual offset, TORIS targets the main cause of HOFs i.e., poor signal quality and propagation condition, by proactively varying the Tx power of the cells whenever a HOF is anticipated. Results show that TORIS outperforms the state-of-the-art HOF reduction solution and yields 40%-75% reduction in HOFs. [ABSTRACT FROM AUTHOR]

We give an overview of recent developments in the modeling of radiowave propagation, based on machine learning (ML) algorithms. We identify the input and output specification and the architecture of the model as the main challenges associated with ML-driven propagation models. Relevant papers are discussed and categorized based on their approach to each of these challenges. Emphasis is given on presenting the prospects and open problems in this promising and rapidly evolving area. [ABSTRACT FROM AUTHOR]

The integrated satellite-terrestrial network (ISN) is an emerging framework that complements the ground 5G network with global network access. The satellite layer consists of Low Earth Orbit (LEO) and Geosynchronous (GEO) satellites for long-distance and high-throughput data transmission with low latency. However, traditional ground routing strategies are not appropriate in ISN, considering the dynamic motion and different capabilities of satellites. Moreover, multi-type tasks need to be allocated with optimal inter-satellite transmission links (ISL) to ensure both the load balance of ISN and the quality of experience (QoE) of subscribers. In this paper, we combine the Software Defined Networking (SDN), AI technique, and fuzzy logic to optimize multi-task routing in ISN. The GEO satellites and the ground computing center (GCC) are set as the joint control plane. The load information of ISN at different times is collected by GEO controllers and made into a multi-D matrix. The GCC collects the historical traffic data from GEO controllers for convolutional neural network (CNN) model training and updating. The GEO satellites utilize the trained CNN model to make path allocation decisions and issue flow tables to LEO satellites. Further, considering the CNN judgment may contradict the user QoE, we use fuzzy logic to evaluate task requirements to improve the CNN output for optimal ISN path allocation. The simulation results demonstrate that the fuzzy-CNN based multi-task routing (FCMR) method has better traffic control performance and path finding flexibility under different ISN conditions. [ABSTRACT FROM AUTHOR]

Traditional Artificial Intelligence (AI) based fault detection approaches need a large amount of data for the model learning. However, in a real-world system, it is very difficult and expensive to obtain massive labeled fault data. In addition, the working conditions of a motor are usually variable, conventional fault diagnosis models with weak generalization ability can only be used for fault detection under constant working condition. The performance of traditional AI based approaches decreases when the working condition changes. To this end, a novel deep Gaussian process (GP) kernel transfer based few-shot learning method (RNGPT) is proposed in this paper for the fault detection of electric machines. First, a deep residual network (ResNet) is used to extract the features of the raw data. Then, the encoded latent feature vector is fed into the GP with kernel transfer ability to make the motor fault detection and classification. The proposed method uses much less data than the traditional AI based method to achieve fault diagnosis under variable working condition, and does not cause an overfitting problem. Experimental results of two case studies demonstrate that the proposed RNGPT model can accurately and effectively detect motor faults with limited labeled data under different working conditions. Experimental results of RNGPT with radial basis function (RBF) kernel model on simulation data present that the fault detection accuracy of the proposed method is about 16% higher than the conventional deep learning methods, 6% higher than other few-shot learning based methods in 5-shot and 4% higher in 1-shot. Finally, experimental on a real-world dataset, the RNGPT-RBF model still has the highest fault diagnosis accuracy in 5-shot (99.39 $ \pm $ 0.09%) and 1-shot (98.55 $ \pm $ 0.16%). [ABSTRACT FROM AUTHOR]

Dissolved gas analysis (DGA) is widely used for oil-immersed power transformers as a conventional fault diagnosis tool. However, interpretation criteria from DGA assessment often depends on empirical discrimination from a specialist, which can render unreliable or ambiguous diagnoses. Intelligent fault classification algorithms can be implemented to conquer uncertainty in conventional methods, and which require feature learning of transformer condition information data rather than expert experience. In this paper, a Gaussian process multi-classification (GPMC) method is proposed, which uses multiclass recognition with a Gaussian process (GP) and renders an output with a probabilistic interpretation rather than a deterministic guess. The method is investigated using large-scale DGA field datasets to improve diagnostic accuracy and presents reliable incipient fault diagnosis ability. A kernel-based learning algorithm and versatile artificial intelligence (AI) methods, (support vector machine (SVM), artificial neural network (ANN), K-nearest neighbors (KNN), decision tree and logistic regression (LR)) have been used to obtain comparative classification accuracy in comparison to the proposed method. Additional comparison is demonstrated between conventional DGA and AI methods. The effectiveness and robustness of the proposed GPMC method are confirmed by experimental accuracy >95%, which illustrates that the proposed method is able to provide superior and reliable diagnoses for operational transformer faults. [ABSTRACT FROM AUTHOR]

The emerging blockchained federated learning, known for its security properties such as decentralization, immutability and traceability, is evolving into an important direction of next-generation AI. With the booming edge computing technologies, blockchained federated learning can take advantage of computing, communication and storage resources geo-distributed at the edge, so that blockchained federated learning can gather edge intelligence from more widely distributed devices more efficiently. However, untrustworthy devices at the edge also bring serious security threats, namely byzantine attacks. Existing solutions focus on selecting local models that are most likely to be honest, rather than detecting byzantine models and identifying attackers, because verifying each local model separately brings intolerable verification delay. In this paper, we propose a byzantine resistant secure blockchained federated learning framework named BytoChain. BytoChain improves the efficiency of model verification by introducing verifiers to execute heavy verification workflows in parallel, and detects byzantine attacks through a byzantine resistant consensus Proof-of-Accuracy (PoA). We analyze how BytoChain can mitigate five types of attacks, and demonstrate its effectiveness by simulations. Finally, we envision some open issues about security, including attacks on privacy, confidentiality, and backdoors. [ABSTRACT FROM AUTHOR]