Machine learning based algorithms for uncertainty quantification in numerical weather prediction models

Moosavi, Azam S. Zavar; Rao, Vishwas; Sandu, Adrian

doi:10.1016/j.jocs.2020.101295

Cited by 41 publications

(15 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…There is no single algorithm that works best under all conditions. [ 45 ]. Firstly, we compared the estimation accuracy of three widely used machine learning models in our study area.…”

Section: Discussionmentioning

confidence: 99%

Machine learning-based estimation of riverine nutrient concentrations and associated uncertainties caused by sampling frequencies

Chen

Zhang

Lin³

et al. 2022

PLoS ONE

View full text Add to dashboard Cite

Accurate and sufficient water quality data is essential for watershed management and sustainability. Machine learning models have shown great potentials for estimating water quality with the development of online sensors. However, accurate estimation is challenging because of uncertainties related to models used and data input. In this study, random forest (RF), support vector machine (SVM), and back-propagation neural network (BPNN) models are developed with three sampling frequency datasets (i.e., 4-hourly, daily, and weekly) and five conventional indicators (i.e., water temperature (WT), hydrogen ion concentration (pH), electrical conductivity (EC), dissolved oxygen (DO), and turbidity (TUR)) as surrogates to individually estimate riverine total phosphorus (TP), total nitrogen (TN), and ammonia nitrogen (NH4+-N) in a small-scale coastal watershed. The results show that the RF model outperforms the SVM and BPNN machine learning models in terms of estimative performance, which explains much of the variation in TP (79 ± 1.3%), TN (84 ± 0.9%), and NH4+-N (75 ± 1.3%), when using the 4-hourly sampling frequency dataset. The higher sampling frequency would help the RF obtain a significantly better performance for the three nutrient estimation measures (4-hourly > daily > weekly) for R2 and NSE values. WT, EC, and TUR were the three key input indicators for nutrient estimations in RF. Our study highlights the importance of high-frequency data as input to machine learning model development. The RF model is shown to be viable for riverine nutrient estimation in small-scale watersheds of important local water security.

show abstract

“…There is no single algorithm that works best under all conditions. [ 45 ]. Firstly, we compared the estimation accuracy of three widely used machine learning models in our study area.…”

Section: Discussionmentioning

confidence: 99%

Machine learning-based estimation of riverine nutrient concentrations and associated uncertainties caused by sampling frequencies

Chen

Zhang

Lin³

et al. 2022

PLoS ONE

View full text Add to dashboard Cite

show abstract

“…Scientists, while running some long or short-range experiments, usually do it, and compare verification results over different configurations. A machine learning-based approach for this process has been proposed in the literature [35,36]. Various microphysics schemes, cumulus parameterizations, and shortwave and longwave radiation schemes were examined, and based on the relationship between the choice of physical processes and the resulting forecast errors, a machine learning model was built to assess WRF model uncertainty.…”

Section: Atmospheric Physics and Processesmentioning

confidence: 99%

Machine Learning in Weather Prediction and Climate Analyses—Applications and Perspectives

2022

View full text Add to dashboard Cite

In this paper, we performed an analysis of the 500 most relevant scientific articles published since 2018, concerning machine learning methods in the field of climate and numerical weather prediction using the Google Scholar search engine. The most common topics of interest in the abstracts were identified, and some of them examined in detail: in numerical weather prediction research—photovoltaic and wind energy, atmospheric physics and processes; in climate research—parametrizations, extreme events, and climate change. With the created database, it was also possible to extract the most commonly examined meteorological fields (wind, precipitation, temperature, pressure, and radiation), methods (Deep Learning, Random Forest, Artificial Neural Networks, Support Vector Machine, and XGBoost), and countries (China, USA, Australia, India, and Germany) in these topics. Performing critical reviews of the literature, authors are trying to predict the future research direction of these fields, with the main conclusion being that machine learning methods will be a key feature in future weather forecasting.

show abstract

“…It also suggests that the thresholds are prone to fluctuation, even when applied conservatively, resulting in false negatives (i.e., bots being mistaken for people) and false positives (i.e., humans being classed as bots). Most social science studies that employ the technology will incorrectly count a significant number of human users as bots, and vice versa [16][17][18][19][20][21][22][23][24][25][26]. There is more research on revealing the communities of such profiles on social media applications [27,28].…”

Section: Related Workmentioning

confidence: 99%

“…If the value of the bootstrap condition is true or in default, the sub-sample size is regulated by the maximum number of samples argument; otherwise, the entire dataset is utilised to create each tree. Being an ensemble learning technique, in most of the binary as well as binary classification problems, it provides reliable solutions with good accuracy [16]. Hence, we have considered it as an existing counterpart to compare the results of our proposed model.…”

Section: Random Forest Classifiermentioning

confidence: 99%

TweezBot: An AI-Driven Online Media Bot Identification Algorithm for Twitter Social Networks

2022

View full text Add to dashboard Cite

In the ultra-connected age of information, online social media platforms have become an indispensable part of our daily routines. Recently, this online public space is getting largely occupied by suspicious and manipulative social media bots. Such automated deceptive bots often attempt to distort ground realities and manipulate global trends, thus creating astroturfing attacks on the social media online portals. Moreover, these bots often tend to participate in duplicitous activities, including promotion of hidden agendas and indulgence in biased propagation meant for personal gain or scams. Thus, online bots have eventually become one of the biggest menaces for social media platforms. Therefore, we have proposed an AI-driven social media bot identification framework, namely TweezBot, which can identify fraudulent Twitter bots. The proposed bot detection method analyzes Twitter-specific user profiles having essential profile-centric features and several activity-centric characteristics. We have constructed a set of filtering criteria and devised an exhaustive bag of words for performing language-based processing. In order to substantiate our research, we have performed a comparative study of our model with the existing benchmark classifiers, such as Support Vector Machine, Categorical Naïve Bayes, Bernoulli Naïve Bayes, Multilayer Perceptron, Decision Trees, Random Forest and other automation identifiers.

show abstract

Machine learning based algorithms for uncertainty quantification in numerical weather prediction models

Cited by 41 publications

References 40 publications

Machine learning-based estimation of riverine nutrient concentrations and associated uncertainties caused by sampling frequencies

Machine learning-based estimation of riverine nutrient concentrations and associated uncertainties caused by sampling frequencies

Machine Learning in Weather Prediction and Climate Analyses—Applications and Perspectives

TweezBot: An AI-Driven Online Media Bot Identification Algorithm for Twitter Social Networks

Contact Info

Product

Resources

About