Distribution Bias Aware Collaborative Generative Adversarial Network for Imbalanced Deep Learning in Industrial IoT

Zhou, Xiaokang; Hu, Yiyong; Wu, Jiayi; Liang, Wei; Ma, Jun; Jin, Qun

doi:10.1109/tii.2022.3170149

Cited by 79 publications

(16 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Nowadays, we can utilize the computing resource of the edge computing environment to train deep learning models for high elasticity. To improve the overall performance of multiple deep learning tasks, Gu et al [64,65] propose to arrange the job scheduling by considering the data cache and deep learning tasks [66]. Wu et al [30] utilize the edge computing framework to achieve image enhancement and object detection in a mobile environment efficiently.…”

Section: Edge Computingmentioning

confidence: 99%

RankXGB-Based Enterprise Credit Scoring by Electricity Consumption in Edge Computing Environment

Shen¹,

Zhang²,

Song³

2023

Computers, Materials &Amp; Continua

View full text Add to dashboard Cite

With the rapid development of the internet of things (IoT), electricity consumption data can be captured and recorded in the IoT cloud center. This provides a credible data source for enterprise credit scoring, which is one of the most vital elements during the financial decision-making process. Accordingly, this paper proposes to use deep learning to train an enterprise credit scoring model by inputting the electricity consumption data. Instead of predicting the credit rating, our method can generate an absolute credit score by a novel deep ranking model-ranking extreme gradient boosting net (rankXGB). To boost the performance, the rankXGB model combines several weak ranking models into a strong model. Due to the high computational cost and the vast amounts of data, we design an edge computing framework to reduce the latency of enterprise credit evaluation. Specially, we design a two-stage deep learning task architecture, including a cloud-based weak credit ranking and an edge-based credit score calculation. In the first stage, we send the electricity consumption data of the evaluated enterprise to the computing cloud server, where multiple weak-ranking networks are executed in parallel to produce multiple weak-ranking results. In the second stage, the edge device fuses multiple ranking results generated in the cloud server to produce a more reliable ranking result, which is used to calculate an absolute credit score by score normalization. The experiments demonstrate that our method can achieve accurate enterprise credit evaluation quickly.

show abstract

Section: Edge Computingmentioning

confidence: 99%

RankXGB-Based Enterprise Credit Scoring by Electricity Consumption in Edge Computing Environment

Shen¹,

Zhang²,

Song³

2023

Computers, Materials &Amp; Continua

View full text Add to dashboard Cite

show abstract

“…The input data is distinguished as either derived from real data or generated data from the generator network, and the differentiation result is fed back to the generator network. Consequently, the generator network adjusts its performance based on the feedback provided by the discriminator network, thereby generating data that more closely resembles real data [27].…”

Section: Discriminatormentioning

confidence: 99%

Cardiac disease diagnosis based on GAN in case of missing data

Chen,

Zhang,

Yang

et al. 2023

Preprint

View full text Add to dashboard Cite

In daily life, two common algorithms are used for collecting medical disease data: data integration of medical institutions and questionnaires. However, these statistical methods require collecting data from the entire research area, which consumes a significant amount of manpower and material resources. Additionally, data integration is difficult and poses privacy protection challenges, resulting in a large number of missing data in the dataset. The presence of incomplete data significantly reduces the quality of the published data, hindering the timely analysis of data and the generation of reliable knowledge by epidemiologists, public health authorities, and researchers.Consequently, this affects the downstream tasks that rely on this data. To address the issue of discrete missing data in cardiac disease, this paper proposes the AGAN(Attribute Generative Adversarial Nets) architecture for missing data filling, based ongenerative adversarial networks. This algorithm takes advantage of the strong learning ability of generative adversarial networks. Given the ambiguous meaning of filling data in other network structures, the attribute matrix is designed to directly convert it into the corresponding data type, making the actual meaning of the filling data more evident. Furthermore, the distribution deviation between the generated data and the real data is integrated into the loss function of the generative adversarial networks,improving their training stability and ensuring consistency between the generated data and the real data distribution. This approach establishes the missing data filling mechanism based on the generative adversarial networks, which ensures the rationality of the data distribution while filling the missing data samples. The experimental results demonstrate that compared to other filling algorithms, the data matrix filled by the proposed algorithm in this paper has more evident practical significance, fewer errors, and higher accuracy in downstream classification prediction.

show abstract

“…Here, the health conditions of students are monitored by wearable sensors embedded in various mobile devices (e.g., smart watches, mobile phones, etc.) [42][43][44]. Thus, we can obtain real-time monitoring data (e.g., electrocardiogram) which can be analyzed and clustered to discover the possible patients from all candidate students.…”

Section: Motivationmentioning

confidence: 99%

Private anomaly detection of student health conditions based on wearable sensors in mobile cloud computing

et al. 2022

View full text Add to dashboard Cite

With the continuous spread of COVID-19 virus, how to guarantee the healthy living of people especially the students who are of relative weak physique is becoming a key research issue of significant values. Specifically, precise recognition of the anomaly in student health conditions is beneficial to the quick discovery of potential patients. However, there are so many students in each school that the education managers cannot know about the health conditions of students in a real-time manner and accurately recognize the possible anomaly among students quickly. Fortunately, the quick development of mobile cloud computing technologies and wearable sensors has provided a promising way to monitor the real-time health conditions of students and find out the anomalies timely. However, two challenges are present in the above anomaly detection issue. First, the health data monitored by massive wearable sensors are often massive and updated frequently, which probably leads to high sensor-cloud transmission cost for anomaly detection. Second, the health data of students are often sensitive enough, which probably impedes the integration of health data in cloud environment even renders the health data-based anomaly detection infeasible. In view of these challenges, we propose a time-efficient and privacy-aware anomaly detection solution for students with wearable sensors in mobile cloud computing environment. At last, we validate the effectiveness and efficiency of our work via a set of simulated experiments.

show abstract

Distribution Bias Aware Collaborative Generative Adversarial Network for Imbalanced Deep Learning in Industrial IoT

Cited by 79 publications

References 26 publications

RankXGB-Based Enterprise Credit Scoring by Electricity Consumption in Edge Computing Environment

RankXGB-Based Enterprise Credit Scoring by Electricity Consumption in Edge Computing Environment

Cardiac disease diagnosis based on GAN in case of missing data

Private anomaly detection of student health conditions based on wearable sensors in mobile cloud computing

Contact Info

Product

Resources

About