PrecipLSTM: A Meteorological Spatiotemporal LSTM for Precipitation Nowcasting

Ma, Zhifeng; Zhang, Hao; Liu, Jie

doi:10.1109/tgrs.2022.3198222

Cited by 19 publications

(22 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Reca = HcA HcA + JaB (34) False Discovery Rate (FDR): The FDR value is calculated using Equation (35).…”

Section: Performance Measuresmentioning

confidence: 99%

“…RNN 34 : The weighted feature

{Wf}_s

is the input to this phase, RNNs are a class of artificial neural networks specifically designed to model sequential or time‐dependent data. Unlike feed‐forward neural networks, which process individual inputs independently, RNNs can capture temporal dependencies by maintaining an internal memory or hidden state.…”

Section: Implementing the New Detection Model For Ddos Attacks Over T...mentioning

confidence: 99%

See 1 more Smart Citation

Detection of distributed denial of service attack using enhanced adaptive deep dilated ensemble with hybrid meta‐heuristic approach

Aliar,

Gowri,

Abins

2024

Trans Emerging Tel Tech

View full text Add to dashboard Cite

The biggest firms throughout the world now are the ones that offer services in the cloud. One of the top problems for cloud users (CUs) and cloud service providers (CSPs) is the availability of cloud‐based services whenever needed. A distributed denial of service (DDoS) assault has been a significant threat to the security of the system in recent years. DDoS defense and detection of these attacks have become hot topics of research for academia and business. However, most approaches cannot accomplish efficient detection outcomes with few false alarms. As a result, minimizing the consequences of DDoS attacks allows CSPs to offer CUs high‐quality services. In the cloud sector, a collective deep structured algorithm is suggested to identify DDoS attacks successfully. The recommended method contains several stages: data acquisition, pre‐processing, optimal feature detection, and selection. The first step is the acquisition of data using the help of publicly available sources. Further, the input data undergoes preprocessing. Consequently, the optimal weighted feature selection takes place on the preprocessed data, where the optimization is done with the aid of the Hybrid Border Collie and Dragonfly Algorithm (HBCDA). Finally, DDoS attack detection is achieved via a novel method known as adaptive deep dilated ensemble (ADDE), which includes, one‐dimensional convolutional neural network (1DCNN), deep temporal convolutional neural network (DTCNN), recurrent neural network (RNN), and bidirectional long short‐term memory (Bi‐LSTM). For the attainment of optimal results, the parameter tuning is accomplished by using the HBCDA approach. The detection outcome is computed by using the fuzzy ranking mechanism. The validation is done for the suggested method model, and its corresponding findings are validated with conventional techniques. Hence, the suggested approach outperforms the detection performance and ensures more efficiency than traditional approaches.

show abstract

“…Reca = HcA HcA + JaB (34) False Discovery Rate (FDR): The FDR value is calculated using Equation (35).…”

Section: Performance Measuresmentioning

confidence: 99%

“…RNN 34 : The weighted feature

{Wf}_s

Section: Implementing the New Detection Model For Ddos Attacks Over T...mentioning

confidence: 99%

Detection of distributed denial of service attack using enhanced adaptive deep dilated ensemble with hybrid meta‐heuristic approach

Aliar,

Gowri,

Abins

2024

Trans Emerging Tel Tech

View full text Add to dashboard Cite

show abstract

“…Utilized entity pairs to partition sentence structures and achieved CNN based feature extraction for effective parts, improving the extraction efficiency. Z. Ma et al [5]. Compared RNN with reference [6], and found that the classification achieved by reference [6] using CNN had the same effect.…”

Section: Introductionmentioning

confidence: 99%

Research on entity relations extraction based on deep learning

Li,

Yang,

Wang

2024

Second International Conference on Electrical, Electronics, and Information Engineering (EEIE 2023)

View full text Add to dashboard Cite

The Internet has brought strong scientific and technological support to social development. With the hardware development rapidly in information technology field, such as digital devices, the application demand of Internet knowledge services by network users has been highly valued by scholars. Entity relation extraction technology is an important technical support for various search engines and machine response artificial intelligence applications, which can help users extract knowledge from massive text data on the Internet. This article use deep learning to study how to enhance the effectiveness and accuracy of entity relationship extraction. Firstly, the basic structure, algorithm principle and process of the most important basic algorithm in neural network are described. The main ways and types of building in-depth model in entity relationship extraction are discussed. After that, the concept of attention-based entity relationship extraction model is put forward, and the structure and implementation method of the model are analyzed. The scheme of improving the model to promote the accuracy and the learning rate. The experimental results indicate that optimization scheme improves the model learning rate and improves the effectiveness of entity phrase pre-training. Compared with other in-depth learning methods, the optimized in-depth learning model proposed can identify the weight of local features more accurately, thus further improving the judgment of the relationship between entity semantics. The model research results text entity relationship extraction have excellent prospect in the field of Web Text Knowledge Service application.

show abstract

“…Thus, here we incorporated the Long Short‐Term Memory (LSTM) to solve the second sub‐problem. LSTM is a classic and widely‐used DL model that can efficiently capture long‐term dependencies of the input sequence by developing gate mechanisms (Shen et al., 2020; Miao et al., 2020; Sahoo et al., 2019; Ma et al., 2022). Some recent studies have reported on the application of LSTM in catching the distinct seasonal variations of precipitation distribution over the TP (Lu & Liu, 2010).…”

Section: Introductionmentioning

confidence: 99%

A Novel Double Machine Learning Strategy for Producing High‐Precision Multi‐Source Merging Precipitation Estimates Over the Tibetan Plateau

Lyu,

Yong

2024

Water Resources Research

View full text Add to dashboard Cite

Precipitation estimation over the Tibetan Plateau is a critical but challenging task due to sparse gauges and high altitudes. Traditional statistic methods are often insufficient to characterize the nonlinear relationship between different precipitation information, while machine learning techniques, particularly deep learning algorithms, offer a novel and powerful approach to improve the merging accuracy of multi‐source precipitation data by efficiently capturing their spatiotemporal dynamics features. This study introduced a novel strategy called Double Machine Learning (DML), which integrates meteorological information, satellite retrievals, and reanalysis data to produce a high‐precision multi‐source merging precipitation product at 0.1° × 0.1°, daily resolution for the Tibetan Plateau. The quantitative evaluation of DML was accomplished using both auto‐meteorological gauges and independent observations. Statistical scores indicate that the new DML‐based merging product apparently outperforms three widely‐used precipitation datasets (IMERG‐Final, GSMaP‐Gauge and ERA5) over the Tibetan Plateau. The proposed DML strategy effectively integrates the advantages of traditional machine learning and deep learning, significantly enhancing the algorithmic robustness and merging accuracy, particularly at medium‐high rain rates in summer. Furthermore, the contributions of multi‐source inputs to the final merging effect was systematically analyzed. It is found that meteorological information, as an auxiliary variable in DML, plays a crucial role in identifying rainy events and adjusting the bias of precipitation estimates, especially over those ungauged regions. This study affirms the call for improving the multi‐source precipitation estimates by combining different machine learning approaches. The new merging precipitation product reported here is recommended for hydrometeorological users of the Tibetan Plateau science community.

show abstract

PrecipLSTM: A Meteorological Spatiotemporal LSTM for Precipitation Nowcasting

Cited by 19 publications

References 26 publications

Detection of distributed denial of service attack using enhanced adaptive deep dilated ensemble with hybrid meta‐heuristic approach

Detection of distributed denial of service attack using enhanced adaptive deep dilated ensemble with hybrid meta‐heuristic approach

Research on entity relations extraction based on deep learning

A Novel Double Machine Learning Strategy for Producing High‐Precision Multi‐Source Merging Precipitation Estimates Over the Tibetan Plateau

Contact Info

Product

Resources

About