The modern stage of the development of telecommunication systems is characterized by the widespread development of wired and wireless data transmission networks. The growth in the number of users of such networks, the...
详细信息
ISBN:
(纸本)9781450387347
The modern stage of the development of telecommunication systems is characterized by the widespread development of wired and wireless data transmission networks. The growth in the number of users of such networks, the emergence of new multimedia services impose high demands on speed, reliability, and delay time in information processing. One of the factors in increasing the data transfer rate is the on-the-fly compression of information at the link level. The paper describes a method for compressing data in a communication channel using error-correcting BCH codes and a feed-forward neural network autoencoder. This method converts a binary vector of user data into BCH codewords, which are used to train the autoencoder. This ultimately makes it possible to reduce the number of transmitted bits in the communication channel.
Background: Single-cell RNA sequencing (scRNA-seq) unfolds complex transcriptomic datasets into detailed cellular maps. Despite recent success, there is a pressing need for specialized methods tailored towards the fun...
详细信息
Background: Single-cell RNA sequencing (scRNA-seq) unfolds complex transcriptomic datasets into detailed cellular maps. Despite recent success, there is a pressing need for specialized methods tailored towards the functional interpretation of these cellular maps. Findings: Here, we present DrivAER, a machine learning approach for the identification of driving transcriptional programs using autoencoder-based relevance scores. DrivAER scores annotated gene sets on the basis of their relevance to user-specified outcomes such as pseudotemporal ordering or disease status. DrivAER iteratively evaluates the information content of each gene set with respect to the outcome variable using autoencoders. We benchmark our method using extensive simulation analysis as well as comparison to existing methods for functional interpretation of scRNA-seq data. Furthermore, we demonstrate that DrivAER extracts key pathways and transcription factors that regulate complex biological processes from scRNA-seq data. Conclusions: By quantifying the relevance of annotated gene sets with respect to specified outcome variables, DrivAER greatly enhances our ability to understand the underlying molecular mechanisms.
Sensors, wearables, mobile devices, and other Internet of Things (IoT) devices are becoming increasingly integrated into all aspects of our lives. They are capable of gathering enormous amounts of data, such as image ...
详细信息
Sensors, wearables, mobile devices, and other Internet of Things (IoT) devices are becoming increasingly integrated into all aspects of our lives. They are capable of gathering enormous amounts of data, such as image data, which can then be sent to the cloud for processing. However, this results in an increase in network traffic and latency. To overcome these difficulties, edge computing has been proposed as a paradigm for computing that brings processing closer to the location where data is produced. This paper explores the merging of cloud and edge computing for IoT and investigates approaches using machine learning for dimensionality reduction of images on the edge, employing the autoencoder deep learning-based approach and principal component analysis (PCA). The encoded data is then sent to the cloud server, where it is used directly for any machine learning task without significantly impacting the accuracy of the data processed in the cloud. The proposed approach has been evaluated on an object detection task using a set of 4000 images randomly chosen from three datasets: COCO, human detection, and HDA datasets. Results show that a 77% reduction in data did not have a significant impact on the object detection task's accuracy.
Stroke and heart disease, which account for a high percentage of the causes of death amongst the elderly population, can occur suddenly leading to death. Hence, early diagnoses and continuous management are required. ...
详细信息
Stroke and heart disease, which account for a high percentage of the causes of death amongst the elderly population, can occur suddenly leading to death. Hence, early diagnoses and continuous management are required. High-risk diseases should be diagnosed through medical personnel using established medical techniques. However, it is time consuming to decide on a diagnosis or the opinion may differ depending on the medical professional. This study aims to shorten the diagnosis period and provide high accuracy diagnoses by establishing the semi-supervised convolution autoencoder and the U-Net models that can classify aortic atherosclerotic plaque conditions and predict the primary locations for stroke occurrence.
Sensor faults are a common type of failure in heat pump systems, which can seriously affect the normal operation of systems. Self-correction of the sensor fault in the system is crucial. State-of-the-art sensor fault ...
详细信息
Sensor faults are a common type of failure in heat pump systems, which can seriously affect the normal operation of systems. Self-correction of the sensor fault in the system is crucial. State-of-the-art sensor fault correction methods based on data-driven and physical models face challenges, such as the need for co-located sensors, accurate physical models, and a large amount of labeled data, greatly limiting their applicability. This paper proposes using machine learning methods for fault self-correction. Firstly, a data self-correction strategy based on the convolutional autoencoder is introduced. Furthermore, an artificial sample generation strategy is proposed to address the scarcity of sensor fault data for data-driven training of the self-correction model. The results demonstrate that the proposed method effectively self-corrects both single and multiple faults. Simultaneously, thermal fault diagnosis evaluations reveal over 90 % accuracy in corrected data, with a maximum diagnostic improvement of 53.5 %. Furthermore, the study shows that the number of parameters is crucial for effective correction, underscoring that over-constraint is essential for successful self-correction.
We seek to better classify canine behavior for guide dog training predictions. Dog temperament is a major factor in success rates and current training also has a blind spot when the puppies are with puppy raisers, who...
详细信息
ISBN:
(纸本)9798400716560
We seek to better classify canine behavior for guide dog training predictions. Dog temperament is a major factor in success rates and current training also has a blind spot when the puppies are with puppy raisers, who are lesser trained volunteers who socialize puppies up to 15 months old. We have used a custom designed smart collar to collect environmental and behavioral data from each puppy individually going through various parts of the guide dog training. We investigate long short-term memory networks (LSTMs), autoencoders (AE), and kernel principal component analysis (KPCA) as methods to identify canine behavior and use multi-sensor data fusion to find the best subset of sensors with the best at classifying temperament. Standard manifold learning experiments take place in controlled environments and translate poorly to real-world applications. This research aims to bridge this gap using guide dog In For Training (IFT) data, which is from a lesser controlled environment and use it to develop a broader data-pattern-to-behavior dictionary for future real-world canine studies.
The deep learning based trackers can always achieve high tracking precision and strong adaptability in different scenarios. However, due to the fact that the number of the parameter is large and the fine-tuning is cha...
详细信息
The deep learning based trackers can always achieve high tracking precision and strong adaptability in different scenarios. However, due to the fact that the number of the parameter is large and the fine-tuning is challenging, the time complexity is high. In order to improve the efficiency, we proposed a tracker based on fast deep learning through constructing a new network with less redundancy. Based on the theory of deep learning, we proposed a deep neural network to describe essential features of images. Furthermore, fast deep learning can be achieved by restricting the size of network. With the help of GPU, the time complexity of the network training is released to a large extent. Under the framework of particle filter, the proposed method combined the deep learning extractor with an SVM scoring professor to distinguish the target from the background. The condensed network structure reduced the complexity of the model. Compared with some other deep learning based tracker, the proposed method can achieve higher efficiency. The frame rate keeps at 22 frames per second on average. Experiments on an open tracking benchmark demonstrate that both the robustness and the timeliness of the proposed tracker are promising when the appearance of the target changes containing translation, rotation and scale or the interference containing illumination, occlusion and cluttered background. Unfortunately, it is not robust enough when the target moves fast or the motion blur and some similar objects exist.
The generation of handwritten Xibo characters is a key step to explore the secrets of this original text. At the same time, it is also a scientific aid to the current task of rescuing and protecting Xibo characters. B...
详细信息
ISBN:
(纸本)9781450384087
The generation of handwritten Xibo characters is a key step to explore the secrets of this original text. At the same time, it is also a scientific aid to the current task of rescuing and protecting Xibo characters. Based on the peculiarities of the structure of the Xibo characters, the prevailing customs of plagiarism from generation to generation, and the reasons for the difficulty of obtaining them at present, the generation of handwritten Xibo characters is a very challenging task. Based on the development of existing handwritten fonts in the field of machine learning, combined with the characteristics of the collected handwritten Xibo font data set, we propose to use a generative adversarial network to try to generate handwritten Xibo fonts. Try the existing generative adversarial network models, they are uncomfortable with the task of generating handwritten Xibo characters. Therefore, this paper proposes a feature adversarial generative model combined with an autoencoder. Using this model to generate handwritten Xibo fonts, the experimental results show that this model can stably generate various handwritten Xibo font images.
The human visual system proves smart in extracting both global and local features. Can we design a similar way for unsupervised feature learning? In this paper, we propose a novel pooling method within an unsupervised...
详细信息
ISBN:
(纸本)9781467395052
The human visual system proves smart in extracting both global and local features. Can we design a similar way for unsupervised feature learning? In this paper, we propose a novel pooling method within an unsupervised feature learning framework, named Rich and Robust Feature Pooling (R~2FP), to better explore rich and robust representation from sparse feature maps of the input data. Both local and global pooling strategies are further considered to instantiate such a method and intensively studied. The former selects the most conductive features in the sub-region and summarizes the joint distribution of the selected features, while the latter is utilized to extract multiple resolutions of features and fuse the features with a feature balancing kernel for rich representation. Extensive experiments on several image recognition tasks demonstrate the superiority of the proposed techniques.
Predictive maintenance uses improved Fuzzy Analytic Hierarchy Process(FAHP) model to calculate the weight of equipment failure and the comprehensive index information of equipment failure, to identify its deterioratio...
详细信息
Predictive maintenance uses improved Fuzzy Analytic Hierarchy Process(FAHP) model to calculate the weight of equipment failure and the comprehensive index information of equipment failure, to identify its deterioration trend, and to judge its trend, to calculate the comprehensive maintenance threshold, to generate maintenance decision information and to identify the equipment locations that need to be disposed of.
暂无评论