Using principal component analysis to detect outliers in ambient air monitoring studies

Brown, Richard; Goddard, Sharon L.; Brown, Andrew S.

doi:10.1080/03067310903094545

Cited by 4 publications

(3 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Despite the fact that data quality is crucial for the development of thermophysical property models, findability, accessibility, interoperability, and reusability (FAIR) data principles 31 are not yet well-established for the actual data quality assessment. Identifying outliers is important not only for thermophysical property data but also for many application domains such as weather forecasting, 32 air quality control, 33 and cybersecurity. 34 Outlier detection, often also referred to as anomaly detection, 35−39 is an important branch of machine learning.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Using Autonomous Outlier Detection Methods for Thermophysical Property Data

Schnorr,

Kaldi,

Staubach

et al. 2024

J. Chem. Eng. Data

View full text Add to dashboard Cite

The reliability and accuracy of thermophysical property data are of central importance for the development of models that describe these properties. In this work, we compare different autonomous algorithms for identifying the outliers in an existing database. Therefore, the comprehensive database on thermophysical property data for the Lennard-Jones fluid [J. Chem. Inf. Model. 2019, 59, 4248−4265] is used. We focus on homogeneous state property data at given temperature and density for the pressure p, thermal expansion coefficient α, isothermal compressibility β, thermal pressure coefficient γ, internal energy u, isochoric heat capacity c v , isobaric heat capacity c p , Gruneisen coefficient Γ, Joule−Thomson coefficient μ JT , speed of sound w, chemical potential μ, (reduced) Helmholtz energy a = a/T, and its derivatives a nm . A comprehensive comparison of 19 outlier detection methods is carried out, which provides insights into the applicability of generic outlier detection algorithms for thermophysical property data. Different classes of outlier detection algorithms are included in the study, namely, machine learning, distance-based, density-based, statistical, ensemble, and model-informed. Two approaches are used for the method evaluation: in approach (a), the original database (comprising real outliers) is used. In approach (b), synthetic outliers are introduced. The results and findings from both approaches are consistent. Machine learning methods yield in some cases better performance compared to that of the distance-based, density-based, ensemble, and statistical methods. The best performance is obtained from the model-informed method (called MoDOD). The results also provide insights into the nature of the outliers in the Lennard-Jones database.

show abstract

Section: Introductionmentioning

confidence: 99%

“…Identifying outliers is important not only for thermophysical property data but also for many application domains such as weather forecasting, air quality control, and cybersecurity . Outlier detection, often also referred to as anomaly detection, − is an important branch of machine learning.…”

Section: Introductionmentioning

confidence: 99%

Using Autonomous Outlier Detection Methods for Thermophysical Property Data

Schnorr,

Kaldi,

Staubach

et al. 2024

J. Chem. Eng. Data

View full text Add to dashboard Cite

show abstract

“…The use of BaP as a 'marker' PAH is an intuitively valid approach, as PAHs are a family of compounds with highly similar physical-chemical properties, and are emitted from similar sources, so they are expected to be present in similar ratios in ambient air. Indeed, correlations between the measured concentrations of certain PAHs and other pollutants in a variety of environmental matrices have been identified in previous studies 7,8 by use of principal component analysis (PCA) 9,10 and diagnostic ratios. 11 These studies have mainly used these techniques as source apportionment tools, 12,13 and the observed correlations or values of ratios between particular PAHs have been discussed as being diagnostic of certain PAH producing processes.…”

Section: Introductionmentioning

confidence: 99%

Correlations in polycyclic aromatic hydrocarbon (PAH) concentrations in UK ambient air and implications for source apportionment

Brown

2012

J. Environ. Monit.

Self Cite

View full text Add to dashboard Cite

The extent of correlation in polycyclic aromatic hydrocarbon (PAH) concentration data obtained by the UK PAH Monitoring and Analysis Air Quality Network from March 2008 to November 2010 has been assessed. Application of principal component analysis (PCA) to the dataset has revealed that the concentrations of the vast majority of PAHs are very highly correlated. The use of diagnostic PAH ratios (including a new benzo[b]naph[2,1-d]thiophene/benzo[a]pyrene diagnostic ratio), and a novel 'combined diagnostic ratio - PCA' approach has revealed information about the main sources of PAH at individual Network sites, allowing the sites to be grouped in terms of those influenced by solid fuel use, industry and traffic, and those of an urban or rural nature. Solid fuel use has also been delineated from other fuel burning. Conclusions are drawn about a number of UK PAH Network sites - four sites are shown to be influenced significantly by solid fuel use, and the sites in South Wales and at London Marylebone Road are found to be located in distinct and unusual PAH pollution climates - the identification of a unique PAH pollution climate in South Wales is a key and novel conclusion of the work.

show abstract

Statistical Methods in Analytical Chemistry

Dementiev

2023

Advances in Geochemistry, Analytical Chemistry, and Planetary Sciences

View full text Add to dashboard Cite

Using principal component analysis to detect outliers in ambient air monitoring studies

Cited by 4 publications

References 12 publications

Using Autonomous Outlier Detection Methods for Thermophysical Property Data

Using Autonomous Outlier Detection Methods for Thermophysical Property Data

Correlations in polycyclic aromatic hydrocarbon (PAH) concentrations in UK ambient air and implications for source apportionment

Statistical Methods in Analytical Chemistry

Contact Info

Product

Resources

About