A neural network aerosol-typing algorithm based on lidar data

Abstract. Atmospheric aerosols play a crucial role in the Earth's system, but their role is not completely understood, partly because of the large variability in their properties resulting from a large number of possible aerosol sources. Recently developed lidar-based techniques were able to retrieve the height distributions of optical and microphysical properties of fine-mode and coarse-mode particles, providing the types of the aerosols. One such technique is based on artificial neural networks (ANNs). In this article, a Neural Network Aerosol Typing Algorithm Based on Lidar Data (NATALI) was developed to estimate the most probable aerosol type from a set of multispectral lidar data. The algorithm was adjusted to run on the EARLINET 3β+2α(+1δ) profiles. The NATALI algorithm is based on the ability of specialized ANNs to resolve the overlapping values of the intensive optical parameters, calculated for each identified layer in the multiwavelength Raman lidar profiles. The ANNs were trained using synthetic data, for which a new aerosol model was developed. Two parallel typing schemes were implemented in order to accommodate data sets containing (or not) the measured linear particle depolarization ratios (LPDRs): (a) identification of 14 aerosol mixtures (high-resolution typing) if the LPDR is available in the input data files, and (b) identification of five predominant aerosol types (low-resolution typing) if the LPDR is not provided. For each scheme, three ANNs were run simultaneously, and a voting procedure selects the most probable aerosol type. The whole algorithm has been integrated into a Python application. The limitation of NATALI is that the results are strongly dependent on the input data, and thus the outputs should be understood accordingly. Additional applications of NATALI are feasible, e.g. testing the quality of the optical data and identifying incorrect calibration or insufficient cloud screening. Blind tests on EARLINET data samples showed the capability of NATALI to retrieve the aerosol type from a large variety of data, with different levels of quality and physical content.

Download Full-text

A Neural Network Aerosol Typing Algorithm Based on Lidar Data

10.5194/acp-2018-492 ◽

2018 ◽

Author(s):

Doina Nicolae ◽

Jeni Vasilescu ◽

Camelia Talianu ◽

Ioannis Binietoglou ◽

Victor Nicolae ◽

...

Keyword(s):

Neural Network ◽

Input Data ◽

Synthetic Data ◽

Optical Data ◽

Lidar Data ◽

Optical Parameters ◽

Physical Content ◽

Large Variability ◽

Aerosol Model ◽

Aerosol Type

Abstract. Atmospheric aerosols play a crucial role in the earth system, but their role is not completely understood, partly because of the large variability in their properties resulting from a large number of possible aerosol sources. Recently developed techniques were able to retrieve the height distributions of optical and microphysical properties of fine-mode and coarse-mode particles, providing the types of the aerosols depicted. One such technique for aerosol typing is based on artificial neural networks (ANNs). In this article, a Neural Network Aerosol Typing Algorithm Based on Lidar Data (NATALI) was developed to estimate the most probable aerosol type from a set of multispectral data. The algorithm has been adjusted for running on the EARLINET 3ß + 2a (+1d) profiles. The NATALI algorithm is based on the ability of specialized ANNs to resolve the overlapping values of the intensive optical parameters calculated for each identified layer in the multiwavelength Raman lidar profiles. The ANNs were trained using synthetic data, for which a new aerosol model was developed. Two parallel typing schemes were implemented in order to accommodate datasets containing or not the measured linear particle depolarization ratios (LPDR): a) identification of mixtures from 14 aerosol mixtures (high-resolution typing) if the LPDR is available in the input data files, and b) identification of 5 predominant aerosol types (low-resolution typing) if the LPDR is not provided. For each scheme, three ANNs were run simultaneously, and a voting procedure selects the most probable answer. The whole algorithm has been integrated into a Python code. The main issue with the approached used in NATALI is that the results are strongly dependent on the input data, and thus the outputs should be understood accordingly. The algorithm has side-applications, for example, to test the quality of the optical data and identify incorrect calibration or incorrect cloud screening. Blind tests on EARLINET data samples showed the capability of this tool to retrieve the aerosol type from a large variety of data, with different quality and physical content.

Download Full-text

Strengths and limitations of the NATALI code for aerosol typing from multiwavelength Raman lidar observations

EPJ Web of Conferences ◽

10.1051/epjconf/201817605005 ◽

2018 ◽

Vol 176 ◽

pp. 05005 ◽

Cited By ~ 1

Author(s):

Doina Nicolae ◽

Camelia Talianu ◽

Jeni Vasilescu ◽

Victor Nicolae ◽

Iwona S. Stachlewska

Keyword(s):

Neural Networks ◽

Optical Parameters ◽

Raman Lidar ◽

Extinction Data ◽

Aerosol Model ◽

Predominant Component ◽

Lidar Observations ◽

Aerosol Type ◽

Made In

A Python code was developed to automatically retrieve the aerosol type (and its predominant component in the mixture) from EARLINET’s 3 backscatter and 2 extinction data. The typing relies on Artificial Neural Networks which are trained to identify the most probable aerosol type from a set of mean-layer intensive optical parameters. This paper presents the use and limitations of the code with respect to the quality of the inputed lidar profiles, as well as with the assumptions made in the aerosol model.

Download Full-text

A Frequency Pattern Mining Model Based on Deep Neural Network for Real-Time Classification of Heart Conditions

Healthcare ◽

10.3390/healthcare8030234 ◽

2020 ◽

Vol 8 (3) ◽

pp. 234 ◽

Cited By ~ 3

Author(s):

Hyun Yoo ◽

Soyoung Han ◽

Kyungyong Chung

Keyword(s):

Neural Network ◽

Big Data ◽

Fourier Transform ◽

Fast Fourier Transform ◽

Real Time ◽

Normal Control ◽

Input Data ◽

Deep Neural Network ◽

Pattern Mining ◽

F Measure

Recently, a massive amount of big data of bioinformation is collected by sensor-based IoT devices. The collected data are also classified into different types of health big data in various techniques. A personalized analysis technique is a basis for judging the risk factors of personal cardiovascular disorders in real-time. The objective of this paper is to provide the model for the personalized heart condition classification in combination with the fast and effective preprocessing technique and deep neural network in order to process the real-time accumulated biosensor input data. The model can be useful to learn input data and develop an approximation function, and it can help users recognize risk situations. For the analysis of the pulse frequency, a fast Fourier transform is applied in preprocessing work. With the use of the frequency-by-frequency ratio data of the extracted power spectrum, data reduction is performed. To analyze the meanings of preprocessed data, a neural network algorithm is applied. In particular, a deep neural network is used to analyze and evaluate linear data. A deep neural network can make multiple layers and can establish an operation model of nodes with the use of gradient descent. The completed model was trained by classifying the ECG signals collected in advance into normal, control, and noise groups. Thereafter, the ECG signal input in real time through the trained deep neural network system was classified into normal, control, and noise. To evaluate the performance of the proposed model, this study utilized a ratio of data operation cost reduction and F-measure. As a result, with the use of fast Fourier transform and cumulative frequency percentage, the size of ECG reduced to 1:32. According to the analysis on the F-measure of the deep neural network, the model had 83.83% accuracy. Given the results, the modified deep neural network technique can reduce the size of big data in terms of computing work, and it is an effective system to reduce operation time.

Download Full-text

Freely scalable and reconfigurable optical hardware for deep learning

Scientific Reports ◽

10.1038/s41598-021-82543-3 ◽

2021 ◽

Vol 11 (1) ◽

Author(s):

Liane Bernstein ◽

Alexander Sludds ◽

Ryan Hamerly ◽

Vivienne Sze ◽

Joel Emer ◽

...

Keyword(s):

Neural Network ◽

Energy Consumption ◽

Data Transfer ◽

Power Delivery ◽

Optical Data ◽

Optical Multicast ◽

Optical Neural Network ◽

Fully Connected ◽

Management Power

AbstractAs deep neural network (DNN) models grow ever-larger, they can achieve higher accuracy and solve more complex problems. This trend has been enabled by an increase in available compute power; however, efforts to continue to scale electronic processors are impeded by the costs of communication, thermal management, power delivery and clocking. To improve scalability, we propose a digital optical neural network (DONN) with intralayer optical interconnects and reconfigurable input values. The path-length-independence of optical energy consumption enables information locality between a transmitter and a large number of arbitrarily arranged receivers, which allows greater flexibility in architecture design to circumvent scaling limitations. In a proof-of-concept experiment, we demonstrate optical multicast in the classification of 500 MNIST images with a 3-layer, fully-connected network. We also analyze the energy consumption of the DONN and find that digital optical data transfer is beneficial over electronics when the spacing of computational units is on the order of $$>10\,\upmu $$ > 10 μ m.

Download Full-text

Comparison of Forecasting Models for Real-Time Monitoring of Water Quality Parameters Based on Hybrid Deep Learning Neural Networks

Water ◽

10.3390/w13111547 ◽

2021 ◽

Vol 13 (11) ◽

pp. 1547

Author(s):

Jian Sha ◽

Xue Li ◽

Man Zhang ◽

Zhong-Liang Wang

Keyword(s):

Neural Network ◽

Water Quality ◽

Deep Learning ◽

Real Time ◽

Best Management Practices ◽

Input Data ◽

Management Practices ◽

Quality Parameters ◽

Water Quality Parameters ◽

Real Time Monitoring

Accurate real-time water quality prediction is of great significance for local environmental managers to deal with upcoming events and emergencies to develop best management practices. In this study, the performances in real-time water quality forecasting based on different deep learning (DL) models with different input data pre-processing methods were compared. There were three popular DL models concerned, including the convolutional neural network (CNN), long short-term memory neural network (LSTM), and hybrid CNN–LSTM. Two types of input data were applied, including the original one-dimensional time series and the two-dimensional grey image based on the complete ensemble empirical mode decomposition algorithm with adaptive noise (CEEMDAN) decomposition. Each type of input data was used in each DL model to forecast the real-time monitoring water quality parameters of dissolved oxygen (DO) and total nitrogen (TN). The results showed that (1) the performances of CNN–LSTM were superior to the standalone model CNN and LSTM; (2) the models used CEEMDAN-based input data performed much better than the models used the original input data, while the improvements for non-periodic parameter TN were much greater than that for periodic parameter DO; and (3) the model accuracies gradually decreased with the increase of prediction steps, while the original input data decayed faster than the CEEMDAN-based input data and the non-periodic parameter TN decayed faster than the periodic parameter DO. Overall, the input data preprocessed by the CEEMDAN method could effectively improve the forecasting performances of deep learning models, and this improvement was especially significant for non-periodic parameters of TN.

Download Full-text

Robust Approach to Supervised Deep Neural Network Training for Real-Time Object Classification in Cluttered Indoor Environment

Applied Sciences ◽

10.3390/app11157148 ◽

2021 ◽

Vol 11 (15) ◽

pp. 7148

Author(s):

Bedada Endale ◽

Abera Tullu ◽

Hayoung Shi ◽

Beom-Soo Kang

Keyword(s):

Neural Network ◽

Deep Learning ◽

Real Time ◽

Network Architecture ◽

Input Data ◽

Deep Neural Network ◽

Data Augmentation ◽

Object Classification ◽

Training Data ◽

Gradient Descent Algorithm

Unmanned aerial vehicles (UAVs) are being widely utilized for various missions: in both civilian and military sectors. Many of these missions demand UAVs to acquire artificial intelligence about the environments they are navigating in. This perception can be realized by training a computing machine to classify objects in the environment. One of the well known machine training approaches is supervised deep learning, which enables a machine to classify objects. However, supervised deep learning comes with huge sacrifice in terms of time and computational resources. Collecting big input data, pre-training processes, such as labeling training data, and the need for a high performance computer for training are some of the challenges that supervised deep learning poses. To address these setbacks, this study proposes mission specific input data augmentation techniques and the design of light-weight deep neural network architecture that is capable of real-time object classification. Semi-direct visual odometry (SVO) data of augmented images are used to train the network for object classification. Ten classes of 10,000 different images in each class were used as input data where 80% were for training the network and the remaining 20% were used for network validation. For the optimization of the designed deep neural network, a sequential gradient descent algorithm was implemented. This algorithm has the advantage of handling redundancy in the data more efficiently than other algorithms.

Download Full-text

Optimal Artificial Neural Network Type Selection Method for Usage in Smart House Systems

Sensors ◽

10.3390/s21010047 ◽

2020 ◽

Vol 21 (1) ◽

pp. 47

Author(s):

Vasyl Teslyuk ◽

Artem Kazarian ◽

Natalia Kryvinska ◽

Ivan Tsmots

Keyword(s):

Neural Network ◽

Neural Networks ◽

Artificial Neural Network ◽

Artificial Neural Networks ◽

Input Data ◽

Optimization Criterion ◽

Fuzzy Input ◽

Different Types ◽

Smart House ◽

Artificial Neural

In the process of the “smart” house systems work, there is a need to process fuzzy input data. The models based on the artificial neural networks are used to process fuzzy input data from the sensors. However, each artificial neural network has a certain advantage and, with a different accuracy, allows one to process different types of data and generate control signals. To solve this problem, a method of choosing the optimal type of artificial neural network has been proposed. It is based on solving an optimization problem, where the optimization criterion is an error of a certain type of artificial neural network determined to control the corresponding subsystem of a “smart” house. In the process of learning different types of artificial neural networks, the same historical input data are used. The research presents the dependencies between the types of neural networks, the number of inner layers of the artificial neural network, the number of neurons on each inner layer, the error of the settings parameters calculation of the relative expected results.

Download Full-text

A New Method of Rolling Prediction for Gas Emission Based on Wavelet Neural Network

Advanced Materials Research ◽

10.4028/www.scientific.net/amr.383-390.1500 ◽

2011 ◽

Vol 383-390 ◽

pp. 1500-1506

Author(s):

Yu Min Pan ◽

Xiao Yu Zhang ◽

Peng Qian Xue

Keyword(s):

Neural Network ◽

Input Data ◽

Wavelet Neural Network ◽

New Method ◽

Neutral Network ◽

Gas Emission ◽

Sample Data ◽

Simulation Results

A new method of rolling prediction for gas emission based on wavelet neural network is proposed in this paper. In the method, part of the sample data is selected, which length is constant, and the data is reselected as the next prediction step. Then a wavelet neutral network is adopted to prediction which input data is rolling, the sequence model of rolling prediction is thus constructed. Simulation results have proved that the method is valid and feasible.

Download Full-text

Artificial neural network-based seismic detector

Bulletin of the Seismological Society of America ◽

10.1785/bssa0850010308 ◽

1995 ◽

Vol 85 (1) ◽

pp. 308-319 ◽

Cited By ~ 6

Author(s):

Jin Wang ◽

Ta-Liang Teng

Keyword(s):

Neural Network ◽

Artificial Neural Network ◽

Input Data ◽

Seismic Event ◽

Signal To Noise Ratio ◽

Type B ◽

Earthquake Detection ◽

Seismic Event Detection ◽

Artificial Neural ◽

Better Than

Abstract An artificial neural network-based pattern classification system is applied to seismic event detection. We have designed two types of Artificial Neural Detector (AND) for real-time earthquake detection. Type A artificial neural detector (AND-A) uses the recursive STA/LTA time series as input data, and type B (AND-B) uses moving window spectrograms as input data to detect earthquake signals. The two AND's are trained under supervised learning by using a set of seismic recordings, and then the trained AND's are applied to another set of recordings for testing. Results show that the accuracy of the artificial neural network-based seismic detectors is better than that of the conventional algorithms solely based on the STA/LTA threshold. This is especially true for signals with either low signal-to-noise ratio or spikelike noises.

Download Full-text

PATTERN RECOGNITION BASED SPEED FORECASTING METHODOLOGY FOR URBAN TRAFFIC NETWORK

Transport ◽

10.3846/16484142.2017.1352027 ◽

2018 ◽

Vol 33 (4) ◽

pp. 959-970 ◽

Cited By ~ 3

Author(s):

Tamás Tettamanti ◽

Alfréd Csikós ◽

Krisztián Balázs Kis ◽

Zsolt János Viharos ◽

István Varga

Keyword(s):

Neural Network ◽

Input Data ◽

Road Traffic ◽

Urban Traffic ◽

Traffic Network ◽

Data Handling ◽

Urban Road ◽

Traffic Sensors ◽

Network Training ◽

Artificial Neural Network Ann

A full methodology of short-term traffic prediction is proposed for urban road traffic network via Artificial Neural Network (ANN). The goal of the forecasting is to provide speed estimation forward by 5, 15 and 30 min. Unlike similar research results in this field, the investigated method aims to predict traffic speed for signalized urban road links and not for highway or arterial roads. The methodology contains an efficient feature selection algorithm in order to determine the appropriate input parameters required for neural network training. As another contribution of the paper, a built-in incomplete data handling is provided as input data (originating from traffic sensors or Floating Car Data (FCD)) might be absent or biased in practice. Therefore, input data handling can assure a robust operation of speed forecasting also in case of missing data. The proposed algorithm is trained, tested and analysed in a test network built-up in a microscopic traffic simulator by using daily course of real-world traffic.

Download Full-text