Drought forecasting using new advanced ensemble-based models of reduced error pruning tree

Shahdad, Mojtaba; Saber, Behzad

doi:10.1007/s11600-022-00738-2

Cited by 18 publications

(8 citation statements)

References 89 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Understanding meteorological variables i.e., trend, variability and future forecasting is essential for comprehending the complexities attributed to climate change. Several ensemble bagging-based models have been utilized for forecasting flood and susceptibility zonation (Chen et al2019;Talukdar et al 2020;Sharon et al 2022), drought (Shahdad and Saber 2022), deforestation prediction (Saha et al 2020), modelling carbon monoxide concentration in the atmosphere (Masih et al 2018) and even rainfall prediction (Bushara and Abraham 2015). However, works on dynamic relationship between rainfall and temperature using ensemble bagging-REPTree model are scant in the existing literature.…”

Section: Introductionmentioning

confidence: 99%

WITHDRAWN: Exploring trend and forecasting of temperature and rainfall in Shimla district of Himachal Pradesh, India using non-parametric and bagging REPTree model

Sharma

Sahana

Saha

et al. 2023

Preprint

View full text Add to dashboard Cite

The changing pattern of climate variables has caused extreme weather events and severe disasters especially in mountainous regions. Such events have a detrimental impact on resources, environment and society. Thus, it has become imperative to examine trends and forecast of meteorological variables using scientific modeling approach at micro level. This study makes an attempt to examine trend in temperature and rainfall using Modified Mann–Kendall test and Sen’s slope estimator during 1980–2021. A Bagging-REPTree machine learning model was utilized for forecasting temperature and rainfall trend for the next 20 years (2022–2041) to understand the temporal dynamics in Shimla district of Indian Himalayan state. Correlation coefficient (R), mean squared error (MSE), mean absolute error (MAE), and root mean squared error mean (RMSE) performance were determined to assess effectiveness and precision of the model. The findings revealed that the frequency of intense rainfall in the district has increased, especially during the monsoon season (June–September) during 1980–2021. Annual maximum, minimum, and mean temperatures have exhibited significant variability while annual rainfall has shown a decreasing trend. The forecast analysis revealed significant trend for rainfall during monsoon season and increasing trend in the maximum temperature has been observed during summer and winter seasons. The analysis has provided sufficient evidence of variability and uncertainty in the behavior of meteorological variables. The outcome of the study may help in devising suitable adaptation and mitigation strategies to combat the effect of climate change in the hilly regions. The methodology adopted in the study may help in future progression of the research in different geographical regions of Western Himalayas.

show abstract

Section: Introductionmentioning

confidence: 99%

WITHDRAWN: Exploring trend and forecasting of temperature and rainfall in Shimla district of Himachal Pradesh, India using non-parametric and bagging REPTree model

Sharma

Sahana

Saha

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…The RMSE and NNSE ranges of the MARS model (during test phase) were 0.37-0.54 and 0.78-0.87, respectively. Shahdad and Saber (2022) [58] investigated the efficiency of the standalone Reduced Error Pruning Tree (REPT) model and its integration with Bagging (BA), Additive Regression (AR), Dagging (DA), and Random Committee (RC) for modeling SPI-6 in the Karkheh watershed in Iran. They found that hybrid algorithms improved the modeling capabilities of the standalone REPT algorithm.…”

Section: Resultsmentioning

confidence: 99%

Drought Forecasting: Application of Ensemble and Advanced Machine Learning Approaches

Hukkeri,

Naganna,

Pruthviraja

et al. 2023

IEEE Access

View full text Add to dashboard Cite

Depending on the severity and spatial-temporal variability, droughts can have a wide range of impacts such as crop failure, water shortages, and food insecurity. Accurate and timely forecasting is necessary to mitigate the hazards of extreme weather events, such as droughts, brought on by climate change. A district like Chitradurga in India, which typically receives around 450-600mm of annual rainfall, will require advanced drought mitigation strategies and plans before the onset of the drought. This research focuses on 1-step lead time forecasting of meteorological drought episodes making use of the 6-month Standardised Precipitation Index (SPI-6) as indicator. The fine resolution rainfall data (0.25°× 0.25°) obtained from the Indian Meteorological Department was used to derive the 6-month SPI data of 23 grid stations. The 1-step lead time SPI-6 time series was forecast considering the antecedent SPI-6 time series data as model input. The Mutual Information was used to determine the most relevant input features for drought forecasting. The standard artificial neural network, an advanced machine learning framework -multivariate adaptive regression splines, and the ensemble learning-based CatBoost regression and gradient tree boosting paradigms were employed to forecast drought episodes. Error and efficiency metrics were employed for performance evaluation of the simulated models. The multivariate adaptive regression splines and gradient tree boosting forecasts had slightly higher accuracy and lower error rates than the artificial neural network model, which suggests that they may be more reliable for drought forecasting. The root mean square error and normalized Nash-Sutcliffe efficiency ranges of the multivariate adaptive regression splines model (during test phase) were 0.37-0.54 and 0.78-0.87, respectively. The thematic maps that were created using spatial interpolation of model forecasts from all the stations also confirmed that the district as a whole experienced drought in April 2019.

show abstract

“…Afterwards, a wrapper‐based selection (Kohavi & John, 1997) is performed on each of the 30 subsets created. These wrappers use the genetic algorithm (GA) metaheuristic as search strategy, the accuracy as fitness measure, and the random forest (RF) (Breiman, 2001), J48 (Sahu & Mehtre, 2015), reduce error pruning tree (REPTree) (Shahdad & Saber, 2022), k‐nearest neighbors ( k ‐NN) (Maleki et al, 2021), support vector machine (SVM) (Chauhan et al, 2019), random tree (Geurts et al, 2006), and Bayes Nets (Ben‐Gal, 2007) learning methods. The KNIME Weka nodes were used to implement these wrappers.…”

Section: Methodsmentioning

confidence: 99%

Examining evolutionary scale modeling‐derived different‐dimensional embeddings in the antimicrobial peptide classification through a KNIME workflow

Martínez‐Mauricio,

García‐Jacas,

Cordoves‐Delgado

2024

Protein Science

View full text Add to dashboard Cite

Molecular features play an important role in different bio‐chem‐informatics tasks, such as the Quantitative Structure–Activity Relationships (QSAR) modeling. Several pre‐trained models have been recently created to be used in downstream tasks, either by fine‐tuning a specific model or by extracting features to feed traditional classifiers. In this regard, a new family of Evolutionary Scale Modeling models (termed as ESM‐2 models) was recently introduced, demonstrating outstanding results in protein structure prediction benchmarks. Herein, we studied the usefulness of the different‐dimensional embeddings derived from the ESM‐2 models to classify antimicrobial peptides (AMPs). To this end, we built a KNIME workflow to use the same modeling methodology across experiments in order to guarantee fair analyses. As a result, the 640‐ and 1280‐dimensional embeddings derived from the 30‐ and 33‐layer ESM‐2 models, respectively, are the most valuable since statistically better performances were achieved by the QSAR models built from them. We also fused features of the different ESM‐2 models, and it was concluded that the fusion contributes to getting better QSAR models than using features of a single ESM‐2 model. Frequency studies revealed that only a portion of the ESM‐2 embeddings is valuable for modeling tasks since between 43% and 66% of the features were never used. Comparisons regarding state‐of‐the‐art deep learning (DL) models confirm that when performing methodologically principled studies in the prediction of AMPs, non‐DL based QSAR models yield comparable‐to‐superior performances to DL‐based QSAR models. The developed KNIME workflow is available‐freely at https://github.com/cicese-biocom/classification-QSAR-bioKom. This workflow can be valuable to avoid unfair comparisons regarding new computational methods, as well as to propose new non‐DL based QSAR models.

show abstract

Drought forecasting using new advanced ensemble-based models of reduced error pruning tree

Cited by 18 publications

References 89 publications

WITHDRAWN: Exploring trend and forecasting of temperature and rainfall in Shimla district of Himachal Pradesh, India using non-parametric and bagging REPTree model

WITHDRAWN: Exploring trend and forecasting of temperature and rainfall in Shimla district of Himachal Pradesh, India using non-parametric and bagging REPTree model

Drought Forecasting: Application of Ensemble and Advanced Machine Learning Approaches

Examining evolutionary scale modeling‐derived different‐dimensional embeddings in the antimicrobial peptide classification through a KNIME workflow

Contact Info

Product

Resources

About