Application of k-means and hierarchical clustering techniques for analysis of air pollution: A review (1980–2019)

Govender, Prinolan; Sivakumar, Venkataraman

doi:10.1016/j.apr.2019.09.009

Cited by 325 publications

(161 citation statements)

References 114 publications

Supporting

Mentioning

149

Contrasting

Unclassified

Order By: Relevance

“…Trajectory clustering was performed using two well-established methods: k-means and Ward linkage (Govender and Sivakumar, 2020) in order to confirm the robustness of our predefined source regions. K-means clustering classifies data into k clusters such that the sum of squares per cluster is minimized (Hartigan and Wong, 1979), with the drawback that k must be specified beforehand.…”

Section: Back Trajectory Classificationmentioning

confidence: 99%

“…Ward linkage is a hierarchical clustering method that merges clusters 195 such that the increase in intra-cluster Ward's distance is minimized (Ward, 1963) and has been described as the method that most closely accomplishes the goals of clustering (Tufféry, 2011). More comprehensive descriptions of these clustering methods can be found elsewhere (Govender and Sivakumar, 2020;Pérez et al, 2017). Prior to clustering, a weighted distance matrix was calculated, similar to Taubman et al (2006): (1) normalized trajectory coordinates to give equal weighting to both horizontal and vertical transport; (2) weighted time steps linearly back in time; and (3) 200assigned nearby points (time step < 6 h) zero weighting on the clustering to remove the influence of aircraft position on the clustering.…”

Section: Back Trajectory Classificationmentioning

confidence: 99%

See 1 more Smart Citation

Long-range transport patterns into the tropical northwest Pacific during the CAMP<sup>2</sup>Ex aircraft campaign: chemical composition, size distributions, and the impact of convection

Hilario¹,

Crosbie²,

Shook³

et al. 2020

Preprint

View full text Add to dashboard Cite

Abstract. The tropical Western North Pacific (TWNP) is a receptor for pollution sources throughout Asia and is highly susceptible to climate change, making it imperative to understand long-range transport in this complex aerosol-meteorological environment. Measurements from the NASA Cloud, Aerosol, and Monsoon Processes Philippines Experiment (CAMP2Ex; 24 Aug to 5 Oct 2019) and back trajectories from the National Oceanic and Atmospheric Administration Hybrid Single Particle Lagrangian Integrated Trajectory Model (HYSPLIT) were used to examine transport into the TWNP from the Maritime Continent (MC), Peninsular Southeast Asia (PSEA), East Asia (EA), and West Pacific (WP). A mid-campaign monsoon shift on 20 Sep 2019 led to distinct transport patterns between the southwest monsoon (before 20 Sep) and monsoon transition (after 20 Sep). During the southwest monsoon, long-range transport was a function of southwesterly winds and cyclones over the South China Sea. Low (high) altitude air generally came from MC (PSEA), implying distinct aerosol processing related to convection and perhaps wind shear. The monsoon transition saw transport from EA and WP, driven by Pacific northeasterly winds, continental anticyclones, and cyclones over the East China Sea. Composition of transported air differed by emission source and accumulated precipitation along trajectories (APT) as an indicator of convection. MC air was characterized by biomass burning tracers while major components of EA air pointed to Asian outflow and secondary formation. Convective scavenging of PSEA air was evidenced by considerable vertical differences between aerosol species but not trace gases, as well as notably higher APT and smaller particles than other regions. Finally, we observed a possible wet scavenging mechanism acting on MC air aloft that was not strictly linked to precipitation. These results are important for understanding the transport and processing of air masses with further implications for modeling aerosol lifecycles and guiding international policymaking on public health and climate.

show abstract

Section: Back Trajectory Classificationmentioning

confidence: 99%

Section: Back Trajectory Classificationmentioning

confidence: 99%

Long-range transport patterns into the tropical northwest Pacific during the CAMP<sup>2</sup>Ex aircraft campaign: chemical composition, size distributions, and the impact of convection

Hilario¹,

Crosbie²,

Shook³

et al. 2020

Preprint

View full text Add to dashboard Cite

show abstract

“…Since then, the different studies performed have made use of other techniques such as genetic algorithms 25 , Hierarchical Agglomerative Clustering 26 , k-means 27 or support vector machines as regressors 28 .…”

Section: Introductionmentioning

confidence: 99%

“…A recent study has shown how k-means clustering can be employed to categorize different locations in a big and populated city representing the variability of pollution according to the variables employed for the study 27 . Finally, the use of support vector machines as a regressor has also been reported in some studies 28 , 29 .…”

Section: Introductionmentioning

confidence: 99%

Evolution and forecasting of PM10 concentration at the Port of Gijon (Spain)

Lasheras

Nieto

García–Gonzalo

et al. 2020

Sci Rep

View full text Add to dashboard Cite

The name PM 10 refers to small particles with a diameter of less than 10 microns. The present research analyses different models capable of predicting PM 10 concentration using the previous values of PM 10 , SO 2 , NO, NO 2 , CO and O 3 as input variables. The information for model training uses data from January 2010 to December 2017. The models trained were autoregressive integrated moving average (ARIMA), vector autoregressive moving average (VARMA), multilayer perceptron neural networks (MLP), support vector machines as regressor (SVMR) and multivariate adaptive regression splines. Predictions were performed from 1 to 6 months in advance. The performance of the different models was measured in terms of root mean squared errors (RMSE). For forecasting 1 month ahead, the best results were obtained with the help of a SVMR model of six variables that gave a RMSE of 4.2649, but MLP results were very close, with a RMSE value of 4.3402. In the case of forecasts 6 months in advance, the best results correspond to an MLP model of six variables with a RMSE of 6.0873 followed by a SVMR also with six variables that gave an RMSE result of 6.1010. For forecasts both 1 and 6 months ahead, ARIMA outperformed VARMA models.

show abstract

“…Traditionally, data clustering is an unsupervised learning task, meaning that the number of clusters is unknown, and none of the input data points are labeled. Applications of clustering include image segmentation [3], text mining [4], gene expression analysis [5], air pollution analysis [6], and fault diagnosis [7], to name only a few.…”

Section: Introductionmentioning

confidence: 99%

Weighted Cluster Ensemble Based on Partition Relevance Analysis With Reduction Step

Ilc

2020

IEEE Access

View full text Add to dashboard Cite

Over the last decade, the advent of the cluster ensemble framework has enabled more accurate and robust data analysis than traditional single clustering algorithms. The improved clustering of microarray data has had a particularly strong impact in the fields of genomics and medicine. However, when we bring several ensemble members together to form a consensus, low-quality data partitions can seriously compromise the final solution. One way to overcome this problem is the weighted cluster ensemble approach based on Partition Relevance Analysis (PRA), which uses internal cluster validity indices to evaluate and weight the ensemble members before the fusion. Unfortunately, the selection of appropriate validation indices for given data is far from trivial. In this paper, we propose an additional step in PRA that reduces the size of the committee of cluster validation indices. It does so by eliminating redundant and noisy indices using data dimensionality reduction methods. Our extension works in an unsupervised way, minimizing the amount of user intervention and required expert knowledge. We adapted three conventional consensus functions based on the principle of evidence accumulation to work with PRA weights. We demonstrate the advantages of the proposed reduction step of PRA based on extensive experiments with 25 gene expression and 15 non-genetic real-world datasets, where we compared 15 consensus functions.

show abstract

Application of k-means and hierarchical clustering techniques for analysis of air pollution: A review (1980–2019)

Cited by 325 publications

References 114 publications

Long-range transport patterns into the tropical northwest Pacific during the CAMP<sup>2</sup>Ex aircraft campaign: chemical composition, size distributions, and the impact of convection

Long-range transport patterns into the tropical northwest Pacific during the CAMP<sup>2</sup>Ex aircraft campaign: chemical composition, size distributions, and the impact of convection

Evolution and forecasting of PM10 concentration at the Port of Gijon (Spain)

Weighted Cluster Ensemble Based on Partition Relevance Analysis With Reduction Step

Contact Info

Product

Resources

About

Application of k-means and hierarchical clustering techniques for analysis of air pollution: A review (1980–2019)

Cited by 325 publications

References 114 publications

Long-range transport patterns into the tropical northwest Pacific during the CAMP&lt;sup&gt;2&lt;/sup&gt;Ex aircraft campaign: chemical composition, size distributions, and the impact of convection

Long-range transport patterns into the tropical northwest Pacific during the CAMP&lt;sup&gt;2&lt;/sup&gt;Ex aircraft campaign: chemical composition, size distributions, and the impact of convection

Evolution and forecasting of PM10 concentration at the Port of Gijon (Spain)

Weighted Cluster Ensemble Based on Partition Relevance Analysis With Reduction Step

Contact Info

Product

Resources

About

Long-range transport patterns into the tropical northwest Pacific during the CAMP<sup>2</sup>Ex aircraft campaign: chemical composition, size distributions, and the impact of convection

Long-range transport patterns into the tropical northwest Pacific during the CAMP<sup>2</sup>Ex aircraft campaign: chemical composition, size distributions, and the impact of convection