scholarly journals Efficacy of Feedforward and LSTM Neural Networks at Predicting and Gap Filling Coastal Ocean Timeseries: Oxygen, Nutrients, and Temperature

2021 ◽  
Vol 8 ◽  
Author(s):  
Steefan Contractor ◽  
Moninya Roughan
Keyword(s):  
Neural Networks ◽  
Decision Making ◽  
Short Term Memory ◽  
Data Sets ◽  
Gap Filling ◽  
Diverse Range ◽  
Big Data Applications ◽  
Free Data ◽  
Before And After ◽  
Ocean Environment

Ocean data timeseries are vital for a diverse range of stakeholders (ranging from government, to industry, to academia) to underpin research, support decision making, and identify environmental change. However, continuous monitoring and observation of ocean variables is difficult and expensive. Moreover, since oceans are vast, observations are typically sparse in spatial and temporal resolution. In addition, the hostile ocean environment creates challenges for collecting and maintaining data sets, such as instrument malfunctions and servicing, often resulting in temporal gaps of varying lengths. Neural networks (NN) have proven effective in many diverse big data applications, but few oceanographic applications have been tested using modern frameworks and architectures. Therefore, here we demonstrate a “proof of concept” neural network application using a popular “off-the-shelf” framework called “TensorFlow” to predict subsurface ocean variables including dissolved oxygen and nutrient (nitrate, phosphate, and silicate) concentrations, and temperature timeseries and show how these models can be used successfully for gap filling data products. We achieved a final prediction accuracy of over 96% for oxygen and temperature, and mean squared errors (MSE) of 2.63, 0.0099, and 0.78, for nitrates, phosphates, and silicates, respectively. The temperature gap-filling was done with an innovative contextual Long Short-Term Memory (LSTM) NN that uses data before and after the gap as separate feature variables. We also demonstrate the application of a novel dropout based approach to approximate the Bayesian uncertainty of these temperature predictions. This Bayesian uncertainty is represented in the form of 100 monte carlo dropout estimates of the two longest gaps in the temperature timeseries from a model with 25% dropout in the input and recurrent LSTM connections. Throughout the study, we present the NN training process including the tuning of the large number of NN hyperparameters which could pose as a barrier to uptake among researchers and other oceanographic data users. Our models can be scaled up and applied operationally to provide consistent, gap-free data to all data users, thus encouraging data uptake for data-based decision making.

scholarly journals CASM: A Deep-Learning Approach for Identifying Collective Action Events with Text and Image Data from Social Media

2019 ◽  
Vol 49 (1) ◽  
pp. 1-57 ◽  
Author(s):  
Han Zhang ◽  
Jennifer Pan
Keyword(s):  
Neural Networks ◽  
Social Media ◽  
Collective Action ◽  
Short Term Memory ◽  
Image Data ◽  
Data Sets ◽  
Event Analysis ◽  
Text Data ◽  
Out Of Sample ◽  
Media Reports

Protest event analysis is an important method for the study of collective action and social movements and typically draws on traditional media reports as the data source. We introduce collective action from social media (CASM)—a system that uses convolutional neural networks on image data and recurrent neural networks with long short-term memory on text data in a two-stage classifier to identify social media posts about offline collective action. We implement CASM on Chinese social media data and identify more than 100,000 collective action events from 2010 to 2017 (CASM-China). We evaluate the performance of CASM through cross-validation, out-of-sample validation, and comparisons with other protest data sets. We assess the effect of online censorship and find it does not substantially limit our identification of events. Compared to other protest data sets, CASM-China identifies relatively more rural, land-related protests and relatively few collective action events related to ethnic and religious conflict.

Improvement of Variant Adaptable LSTM Trained With Metaheuristic Algorithms for Healthcare Analysis

2021 ◽  
pp. 1031-1051
Author(s):  
Tarik A. Rashid ◽  
Mohammad K. Hassan ◽  
Mokhtar Mohammadi ◽  
Kym Fraser
Keyword(s):  
Diabetes Mellitus ◽  
Neural Networks ◽  
Time Series Data ◽  
Short Term Memory ◽  
Learning Algorithm ◽  
Back Propagation ◽  
Machine Learning Techniques ◽  
Series Data ◽  
Data Sets ◽  
Patients With Diabetes

Recently, the population of the world has increased along with health problems. Diabetes mellitus disease as an example causes issues to the health of many patients globally. The task of this chapter is to develop a dynamic and intelligent decision support system for patients with different diseases, and it aims at examining machine-learning techniques supported by optimization techniques. Artificial neural networks have been used in healthcare for several decades. Most research works utilize multilayer layer perceptron (MLP) trained with back propagation (BP) learning algorithm to achieve diabetes mellitus classification. Nonetheless, MLP has some drawbacks, such as, convergence, which can be slow; local minima can affect the training process. It is hard to scale and cannot be used with time series data sets. To overcome these drawbacks, long short-term memory (LSTM) is suggested, which is a more advanced form of recurrent neural networks. In this chapter, adaptable LSTM trained with two optimizing algorithms instead of the back propagation learning algorithm is presented. The optimization algorithms are biogeography-based optimization (BBO) and genetic algorithm (GA). Dataset is collected locally and another benchmark dataset is used as well. Finally, the datasets fed into adaptable models; LSTM with BBO (LSTMBBO) and LSTM with GA (LSTMGA) for classification purposes. The experimental and testing results are compared and they are promising. This system helps physicians and doctors to provide proper health treatment for patients with diabetes mellitus. Details of source code and implementation of our system can be obtained in the following link “https://github.com/hamakamal/LSTM.”

Spatial variability in periglacial terrain conditions, northwestern Canada

2021 ◽  
Author(s):  
Peter Morse ◽  
Wendy Sladen ◽  
Steve Kokelj ◽  
Ryan Parker ◽  
Sharon Smith ◽  
...  
Keyword(s):  
Decision Making ◽  
Systems Approach ◽  
Mass Movement ◽  
Spatial Models ◽  
Data Sets ◽  
Mackenzie Delta ◽  
Land Area ◽  
Diverse Range ◽  
Surficial Geology ◽  
Physiographic Regions

<p>Throughout much of northern Canada there is an inadequate knowledge of permafrost and periglacial terrain conditions, which impedes development of climate-resilient northern infrastructure, identification of potential geohazards, decision making regarding resource development, and inferring past and future landscape evolution. Using a land systems approach to better understand formation of landscapes and thaw-sensitive terrains of northern Yukon and northwestern Northwest Territories, we aim to describe the permafrost-related landform-sediment assemblages that exist in the region. Permafrost is continuous in the region, but variations in geology, landscape history, climate, relief, ecology, and other natural processes have produced a diverse range of permafrost conditions and landforms. Using the 875 km-long Dempster and Inuvik-to-Tuktoyaktuk highway corridors (DH-ITH) as a regional transect, and high-resolution satellite imagery, a robust methodology was implemented to identify and digitize (at 1:5000 scale) 8793 landforms (589 km<sup>2</sup>) within a 10 km-wide corridor (8530 km<sup>2</sup>) and classify them according to main formational process (hydrological, periglacial, and mass movement). Surficial geology data were extracted from available data sets. Landform densities in all feature classes vary substantially along the transect according to physiographic region and surficial geology. The northern 39% of the corridor is characterized by generally rolling or planar relief, numerous waterbodies (19%), and the remaining land area by mostly morainal (67%), glaciofluvial (12%), lacustrine (7%), and alluvial (7%) deposits. By count, it contains 53% of mapped features and the majority of periglacial (67%) and hydrological (70%) features. In particular, the Tuktoyaktuk Coastlands, Peel Plain, and Mackenzie Delta, contain the greatest density of mapped landforms within the corridor, which cover nearly 23%, 15%, and 15% of the land area of these physiographic regions, respectively. These extents reflect the amount of ground ice and level of permafrost-thaw sensitivity of these regions. In contrast, the physiographic regions of the southern 61% of the study area are characterized by high relative relief, few waterbodies (0.2%), and the land area mainly by colluvial (63%), alluvial (18%), and morainal (14%) deposits. Most mass movement features occur here (85% by count), and are concentrated in the Ogilvie Mountains (n = 1027; 108 km<sup>2</sup>). This feature inventory provides the basis for developing spatial models of landscape-thaw susceptibility, which can inform risk assessment and improve decision making regarding public safety and environmental management.</p>

scholarly journals RippleNet: A Recurrent Neural Network for Sharp Wave Ripple (SPW-R) Detection

2020 ◽  
Author(s):  
Espen Hagen ◽  
Anna R. Chambers ◽  
Gaute T. Einevoll ◽  
Klas H. Pettersen ◽  
Rune Enger ◽  
...  
Keyword(s):  
Neural Networks ◽  
Open Source ◽  
Short Term Memory ◽  
Field Potential ◽  
Low Frequency ◽  
Data Sets ◽  
Brain Functions ◽  
Sharp Wave ◽  
Ca1 Region ◽  
Multidisciplinary Study

AbstractHippocampal sharp wave ripples (SPW-R) have been identified as key bio-markers of important brain functions such as memory consolidation and decision making. SPW-R detection typically relies on hand-crafted feature extraction, and laborious manual curation is often required. In this multidisciplinary study, we propose a novel, self-improving artificial intelligence (AI) method in the form of deep Recurrent Neural Networks (RNN) with Long Short-Term memory (LSTM) layers that can learn features of SPW-R events from raw, labeled input data. The algorithm is trained using supervised learning on hand-curated data sets with SPW-R events. The input to the algorithm is the local field potential (LFP), the low-frequency part of extracellularly recorded electric potentials from the CA1 region of the hippocampus. The output prediction can be interpreted as the time-varying probability of SPW-R events for the duration of the input. A simple thresholding applied to the output probabilities is found to identify times of events with high precision. The reference implementation of the algorithm, named ‘RippleNet’, is open source, freely available, and implemented using a common open-source framework for neural networks (tensorflow.keras) and can be easily incorporated into existing data analysis workflows for processing experimental data.

scholarly journals Testing the applicability of neural networks as a gap-filling method using CH<sub>4</sub> flux data from high latitude wetlands

2013 ◽  
Vol 10 (12) ◽  
pp. 8185-8200 ◽  
Author(s):  
S. Dengel ◽  
D. Zona ◽  
T. Sachs ◽  
M. Aurela ◽  
M. Jammet ◽  
...  
Keyword(s):  
Neural Networks ◽  
Time Series ◽  
Pearson Correlation ◽  
Time Of Day ◽  
Gap Filling ◽  
Ch4 Flux ◽  
Advantages And Disadvantages ◽  
Flux Data ◽  
Free Data ◽  
Filling Method

Abstract. Since the advancement in CH4 gas analyser technology and its applicability to eddy covariance flux measurements, monitoring of CH4 emissions is becoming more widespread. In order to accurately determine the greenhouse gas balance, high quality gap-free data is required. Currently there is still no consensus on CH4 gap-filling methods, and methods applied are still study-dependent and often carried out on low resolution, daily data. In the current study, we applied artificial neural networks to six distinctively different CH4 time series from high latitudes, explain the method and test its functionality. We discuss the applicability of neural networks in CH4 flux studies, the advantages and disadvantages of this method, and what information we were able to extract from such models. Three different approaches were tested by including drivers such as air and soil temperature, barometric air pressure, solar radiation, wind direction (indicator of source location) and in addition the lagged effect of water table depth and precipitation. In keeping with the principle of parsimony, we included up to five of these variables traditionally measured at CH4 flux measurement sites. Fuzzy sets were included representing the seasonal change and time of day. High Pearson correlation coefficients (r) of up to 0.97 achieved in the final analysis are indicative for the high performance of neural networks and their applicability as a gap-filling method for CH4 flux data time series. This novel approach which we show to be appropriate for CH4 fluxes is a step towards standardising CH4 gap-filling protocols.

scholarly journals A Survey on Some Big Data Applications Tools and Technologies

2021 ◽  
Vol 9 (6) ◽  
pp. 239-242
Author(s):  
Nazia Tazeen ◽  
Sandhya Rani K.
Keyword(s):  
Decision Making ◽  
Big Data ◽  
Data Transmission ◽  
Data Sets ◽  
Complex Data ◽  
Broad Area ◽  
Apache Hadoop ◽  
Engineering Data ◽  
Big Data Applications ◽  
Science Media

Big Data is a broad area that deals with enormous chunks of data sets. It is a word for enormous data sets having huge volume, more diverse structures of data originating from diverse sources are growing rapidly. Many data being generated because of fast data transmission between devices concerning different sectors like healthcare, science, media, business, entertainment and engineering. Data collection capacity and its storage is big concern. Apache Hadoop software is a store of accessible source programs to store big data and perform analytics and various other operations related to big data. Many organizations base their decisions by extracting knowledge from huge and complex data, because of this prime cause of decision making, Big Data has to be accurately classified and analyzed. In order to overcome the complex challenges encountered by Big Data, various Big Data tools and technologies have developed. Big Data Applications, tools and technologies used to handle it are briefly discussed in this paper.

scholarly journals The Flip-flop neuron – A memory efficient alternative for solving challenging sequence processing and decision making problems

2021 ◽  
Author(s):  
Sweta Kumari ◽  
Vigneswaran C ◽  
V. Srinivasa Chakravarthy
Keyword(s):  
Neural Networks ◽  
Decision Making ◽  
Information Integration ◽  
Short Term Memory ◽  
Previous History ◽  
Video Frame ◽  
Sequential Decision ◽  
Flip Flop ◽  
Sequence Processing ◽  
Difficult Sequence

Sequential decision making tasks that require information integration over extended durations of time are challenging for several reasons including the problem of vanishing gradients, long training times and significant memory requirements. To this end we propose a neuron model fashioned after the JK flip-flops in digital systems. A flip-flop is a sequential device that can store state information of the previous history. We incorporate the JK flip-flop neuron into several deep network architectures and apply the networks to difficult sequence processing problems. The proposed architectures include flip-flop neural networks (FFNNs), bidirectional flip-flop neural networks (BiFFNNs), convolutional flip-flop neural networks (ConvFFNNs), and bidirectional convolutional flip-flop neural networks (BiConvFFNNs). Learning rules of proposed architectures have also been derived. We have considered the most popular benchmark sequential tasks like signal generation, sentiment analysis, handwriting generation, text generation, video frame prediction, lung volume prediction, and action recognition to evaluate the proposed networks. Finally, we compare the results of our networks with the results from analogous networks with Long Short-Term Memory (LSTM) neurons on the same sequential tasks. Our results show that the JK flip-flop networks outperform the LSTM networks significantly or marginally on all the tasks, with only half of the trainable parameters.

Improvement of Variant Adaptable LSTM Trained With Metaheuristic Algorithms for Healthcare Analysis

2019 ◽  
pp. 111-131 ◽  
Author(s):  
Tarik A. Rashid ◽  
Mohammad K. Hassan ◽  
Mokhtar Mohammadi ◽  
Kym Fraser
Keyword(s):  
Diabetes Mellitus ◽  
Neural Networks ◽  
Time Series Data ◽  
Short Term Memory ◽  
Learning Algorithm ◽  
Back Propagation ◽  
Machine Learning Techniques ◽  
Series Data ◽  
Data Sets ◽  
Patients With Diabetes

Recently, the population of the world has increased along with health problems. Diabetes mellitus disease as an example causes issues to the health of many patients globally. The task of this chapter is to develop a dynamic and intelligent decision support system for patients with different diseases, and it aims at examining machine-learning techniques supported by optimization techniques. Artificial neural networks have been used in healthcare for several decades. Most research works utilize multilayer layer perceptron (MLP) trained with back propagation (BP) learning algorithm to achieve diabetes mellitus classification. Nonetheless, MLP has some drawbacks, such as, convergence, which can be slow; local minima can affect the training process. It is hard to scale and cannot be used with time series data sets. To overcome these drawbacks, long short-term memory (LSTM) is suggested, which is a more advanced form of recurrent neural networks. In this chapter, adaptable LSTM trained with two optimizing algorithms instead of the back propagation learning algorithm is presented. The optimization algorithms are biogeography-based optimization (BBO) and genetic algorithm (GA). Dataset is collected locally and another benchmark dataset is used as well. Finally, the datasets fed into adaptable models; LSTM with BBO (LSTMBBO) and LSTM with GA (LSTMGA) for classification purposes. The experimental and testing results are compared and they are promising. This system helps physicians and doctors to provide proper health treatment for patients with diabetes mellitus. Details of source code and implementation of our system can be obtained in the following link “https://github.com/hamakamal/LSTM.”

Sentiment Recognition in Customer Reviews Using Deep Learning

2020 ◽  
pp. 1383-1393
Author(s):  
Vinay Kumar Jain ◽  
Shishir Kumar ◽  
Prabhat Kumar Mahanti
Keyword(s):  
Neural Networks ◽  
Deep Learning ◽  
Sentiment Analysis ◽  
Text Classification ◽  
Short Term Memory ◽  
Machine Learning Techniques ◽  
Data Sets ◽  
Customer Reviews ◽  
Weakly Supervised ◽  
Comparison Of The Results

Deep learning has become popular in all aspect related to human judgments. Most machine learning techniques work well which includes text classification, text sequence learning, sentiment analysis, question-answer engine, etc. This paper has been focused on two objectives, firstly is to study the applicability of deep neural networks strategies for extracting sentiment present in social media data and customer reviews with effective training solutions. The second objective is to design deep networks that can be trained with these weakly supervised strategies in order to predict meaningful inferences. This paper presents the concept and steps of using deep learning for extraction sentiments from customer reviews. The extraction pulls out the features from the customer reviews using deep learning popular methods including Convolution neural networks (CNN) and Long Short-Term Memory (LSTM) architectures. The comparison of the results with tradition text classification method such as Naive Bayes(NB) and Support Vector Machine(SVM) using two data sets IMDB reviews and Amazon customer reviews have been presented. This work mainly focused on investigating the merit of using deep models for sentiment analysis in customer reviews.

scholarly journals Modeling moisture fluxes using artificial neural networks: can information extraction overcome data loss?

2010 ◽  
Vol 7 (5) ◽  
pp. 6525-6551
Author(s):  
A. L. Neal ◽  
H. V. Gupta ◽  
S. A. Kurc ◽  
P. D. Brooks
Keyword(s):  
Neural Networks ◽  
Artificial Neural Networks ◽  
Eddy Covariance ◽  
Land Surface ◽  
Data Sets ◽  
Data Loss ◽  
Gap Filling ◽  
Multiple Data ◽  
Moisture Fluxes ◽  
Artificial Neural

Abstract. Eddy covariance sites can experience data losses as high as 30 to 45% on an annual basis. Artificial neural networks (ANNs) have been identified as powerful tools for gap filling, but their performance depends on the representativeness of data used to train the model. In this paper, we develop a normalization method, which has similar performance compared to conventional training approaches, but exhibits differences in the timing of fluxes, indicating different and previously unused information in the data record. Specifically, the differences between half-hourly model fluxes, especially during summer months, indicate that the structure of the information content in the data changes seasonally, diurnally and with the rate of data loss. This variation between gap-filling models complicates the application of their output as consistent data sets for land surface modeling, and points to the need for improved data and models to address flux behavior at critical times. We advise several approaches to address these concerns, including use of separate models for day and nighttime processes and the use of multiple data streams at dawn, when eddy covariance may be particularly ineffective due to the timing of the onset of turbulent mixing.

Sign in / Sign up

Export Citation Format

Share Document