Hybrid Model of Convolutional LSTM and CNN to Predict Particulate Matter

Lee, Seonggu; Shin, Jitae

doi:10.18178/ijiee.2019.9.1.701

Cited by 10 publications

(6 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…PConvLSTM-P (G) has the same structure as PConvLSTM-P and is trained with the proposed generating missing technique. PConvLSTM-P (G) is optimized with L total defined in Equation 28, whereas the other three networks are trained using L obs and L tv defined in Equation (23) and (24) respectively. We compare ConvLSTM and PConvLSTM-C to examine the performance of the PConvLSTM layer.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Partial Convolutional LSTM for Spatiotemporal Prediction of Incomplete Data

Son

Jang

2020

IEEE Access

View full text Add to dashboard Cite

Advanced data analysis techniques facilitate data-driven spatiotemporal prediction in various fields. However, in real-world data, missing values are inevitable, which causes the data incomplete and makes predictions more challenging. Although we can train complex spatiotemporal correlations with deep learning techniques, most deep learning networks require data without any missing values. In this paper, we propose a novel deep learning framework that manages missing values in the grid-based data structure. We design a partial convolutional long-short-term-memory (PConvLSTM) by combining partial convolution for inpainting and convolutional long-short-term-memory (ConvLSTM) for spatiotemporal prediction. We treat incomplete spatiotemporal data with the partial convolution and train spatiotemporal dependencies with the ConvLSTM structure. The trained PConvLSTM can predict continuous spatial data with missing regions in incomplete input data. We also train the network using incomplete spatiotemporal data without ground truth to enhance practicality. Existing deep learning networks to interpolate missing data are mostly trained by applying ground truth data without missing regions. We show that PConvLSTM achieves higher prediction accuracies compared to ConvLSTM for incomplete data without ground truth.

show abstract

Section: Methodsmentioning

confidence: 99%

“…However, irregularities result in missing regions after mapping the data points on a grid. Lee and Shin [24] have introduced a technique for estimating missing regions with inverse distance weighting interpolation and deep learning training. However, interpolation might produce biased results depending on techniques [25].…”

Section: Related Workmentioning

confidence: 99%

Partial Convolutional LSTM for Spatiotemporal Prediction of Incomplete Data

Son

Jang

2020

IEEE Access

View full text Add to dashboard Cite

show abstract

“…A multivariate deep learning model is constructed to analyze IoT time-series data collected from multiple sources with various domains [17,18]. From existing studies, we can examine the general structure of deep learning models used to analyze multivariate IoT time-series data.…”

Section: Deep Learning-based Multivariate Iot Time-series Analysismentioning

confidence: 99%

Task-Driven Transferred Vertical Federated Deep Learning for Multivariate Internet of Things Time-Series Analysis

Oh,

Lee

2024

Applied Sciences

View full text Add to dashboard Cite

As big data technologies for IoT services develop, cross-service distributed learning techniques of multivariate deep learning models on IoT time-series data collected from various sources are becoming important. Vertical federated deep learning (VFDL) is used for cross-service distributed learning for multivariate IoT time-series deep learning models. Existing VFDL methods with reasonable performance require a large communication amount. On the other hand, existing communication-efficient VFDL methods have relatively low performance. We propose TT-VFDL-SIM, which can achieve improved performance over centralized training or existing VFDL methods in a communication-efficient manner. TT-VFDL-SIM derives partial tasks from the target task and applies transfer learning to them. In our task-driven transfer approach for the design of TT-VFDL-SIM, the SIM Partial Training mechanism contributes to performance improvement by introducing similar feature spaces in various ways. TT-VFDL-SIM was more communication-efficient than existing VFDL methods and achieved an average of 0.00153 improved MSE and 7.98% improved accuracy than centralized training or existing VFDL methods.

show abstract

“…Particulate matter (PM), also recognised as particle pollution, is a term that is used to describe a combination of liquid and solid droplets suspended in the air [1]. PM, which typically used as a measure of air quality, is one of the common measurement metrics among other environment-related parameters such as volatile organic compound (VOC), carbon dioxide (CO 2 ), humidity and temperature [2][3][4][5].…”

Section: Introductionmentioning

confidence: 99%

Laboratory and field evaluation of three low‐cost particulate matter sensors

Ghamari

Soltanpur

Rangel

et al. 2022

IET Wireless Sensor Systems

View full text Add to dashboard Cite

Low‐cost off‐the‐shelf particulate matter (PM) sensors have the potentiality to be used for evaluating the air quality in outdoor settings. Monitoring of air quality in surface coal mines is an example of such applications. In coal mines, long‐term exposure to inhalation of coal dust is harmful and can lead to coal workers' pneumoconiosis, which is a potentially disabling lung disease. Therefore, continual monitoring of air quality in coal mines is a must and vital and can potentially assist in preventing such diseases. Although, using and deploying of the existing low‐cost and lightweight sensors can help to improve monitoring resolution in a much cost‐effective manner, there are some concerns regarding the reliability of the collected data from these sensors. Therefore, low‐cost PM sensors are required to initially be compared with the standard reference instruments and then be calibrated. In this study, three different types of low‐cost, light‐scattering‐based widely available PM sensors (Shinyei PPD42NS, Sharp GP2Y1010AU0F and Laser SEN0177) are evaluated, compared, and calibrated with the reference instruments in a controlled environment as well as in a field experiment (surface coal mine).

show abstract

Hybrid Model of Convolutional LSTM and CNN to Predict Particulate Matter

Cited by 10 publications

References 9 publications

Partial Convolutional LSTM for Spatiotemporal Prediction of Incomplete Data

Partial Convolutional LSTM for Spatiotemporal Prediction of Incomplete Data

Task-Driven Transferred Vertical Federated Deep Learning for Multivariate Internet of Things Time-Series Analysis

Laboratory and field evaluation of three low‐cost particulate matter sensors

Contact Info

Product

Resources

About