Comparison of missing value imputation methods in time series: the case of Turkish meteorological data

Yozgatlıgil, Ceylan; Aslan, Sipan; İyigün, Cem; Batmaz, İnci

doi:10.1007/s00704-012-0723-x

Cited by 146 publications

(114 citation statements)

References 52 publications

Supporting

Mentioning

110

Contrasting

Unclassified

Order By: Relevance

“…In [14] they compared the internal imputation offered by the classifiers C4.5 and CN2 with that of KNN imputation and mean imputation, with KNN being the best. In [19], they applied mean imputation, normal ratio method, normal ratio with correlation method, multilayer perceptron network and multiple imputation to meteorological time series data sets. Multiple imputation, although being the most computationally demanding, was more robust and outperformed the rest.…”

Section: Literature Review Of Imputation Techniquesmentioning

confidence: 99%

Dealing with missing data for prognostic purposes

Loukopoulos

Sampath

Pilidis

et al. 2016

2016 Prognostics and System Health Management Conference (PHM-Chengdu)

View full text Add to dashboard Cite

Abstract-Centrifugal compressors are considered one of the most critical components in oil industry, making the minimisation of their downtime and the maximisation of their availability a major target. Maintenance is thought to be a key aspect towards achieving this goal, leading to various maintenance schemes being proposed over the years. Condition based maintenance and prognostics and health management (CBM/PHM), which is relying on the concepts of diagnostics and prognostics, has been gaining ground over the last years due to its ability of being able to plan the maintenance schedule in advance. The successful application of this policy is heavily dependent on the quality of data used and a major issue affecting it, is that of missing data. Missing data's presence may compromise the information contained within a set, thus having a significant effect on the conclusions that can be drawn from the data, as there might be bias or misleading results. Consequently, it is important to address this matter. A number of methodologies to recover the data, called imputation techniques, have been proposed. This paper reviews the most widely used techniques and presents a case study with the use of actual industrial centrifugal compressor data, in order to identify the most suitable ones.

show abstract

Section: Literature Review Of Imputation Techniquesmentioning

confidence: 99%

Dealing with missing data for prognostic purposes

Loukopoulos

Sampath

Pilidis

et al. 2016

2016 Prognostics and System Health Management Conference (PHM-Chengdu)

View full text Add to dashboard Cite

show abstract

“…The autoregressive integrated moving-average (ARIMA) model [17] and simple exponential smoothing (SES) [18] are two representative examples of this approach. However, this approach fails to address two major problems.…”

Section: Introductionmentioning

confidence: 99%

A Two-Step Method for Missing Spatio-Temporal Data Reconstruction

Cheng

2017

IJGI

View full text Add to dashboard Cite

Abstract:Missing data reconstruction is a critical step in the analysis and mining of spatio-temporal data; however, few studies comprehensively consider missing data patterns, sample selection and spatio-temporal relationships. As a result, traditional methods often fail to obtain satisfactory accuracy or address high levels of complexity. To combat these problems, this study developed an effective two-step method for spatio-temporal missing data reconstruction (ST-2SMR). This approach includes a coarse-grained interpolation method for considering missing patterns, which can successfully eliminate the influence of continuous missing data on the overall results. Based on the results of coarse-grained interpolation, a dynamic sliding window selection algorithm was implemented to determine the most relevant sample data for fine-grained interpolation, considering both spatial and temporal heterogeneity. Finally, spatio-temporal interpolation results were integrated by using a neural network model. We validated our approach using Beijing air quality data and found that the proposed method outperforms existing solutions in term of estimation accuracy and reconstruction rate.

show abstract

“…Young (1992) presenta una modificación al método RN, donde la razón-normal ponderada es cambiada por la correlación entre las estaciones (NRWC) pero Yozgatligil et al (2013) reportan que el promedio aritmético simple (SAA) y (NRWC) son métodos que son idénticos en algunos casos, si las correlaciones de estaciones de referencia son casi la misma.…”

Section: Introductionunclassified

“…Young (1992) introduced an modification to RN method, where reason-normal weighted is changed by the correlation between stations (NRWC) but Yozgatligil et al (2013) report that simple arithmetic average (SAA) and (NRWC) are methods that are identical in some cases, if the correlations of reference stations are almost the same.…”

Section: Introductionmentioning

confidence: 99%

“…McCuen (1998) recommends the simple arithmetic average (SAA) method, when the annual value in each data to average differs at least 10%, but according to Yozgatligil (2013) presents reliable estimates if the variable has no spatial variability and if the reference stations are variable no tiene variabilidad espacial y si las estaciones de referencia están altamente correlacionadas. ASCE (1996), propone el método de ponderación de distancia inversa IDW, también conocido como el método U.S National Weather Servic por su implementación en estudios hidrológicos y geográficos al igual que estudios realizados por Hubbard (1994);Sokol y Stekl (1994); Palomino y Martin (1995); Teegavarapu y Chandramouli (2005).…”

Section: Introductionunclassified

See 1 more Smart Citation

Relleno de series diarias de precipitación, temperatura mínima, máxima de la región norte del Urabá Antioqueño

Trujillo

Arteaga-Ramírez

Vázquez-Peña

et al. 2017

Remexca

View full text Add to dashboard Cite

ResumenLos estudios agroclimáticos requieren la utilización de series cronológicas de las variables meteorológicas, que generalmente presentan datos faltantes limitando su uso, por lo que los métodos de relleno son utilizados. La finalidad de este trabajo fue determinar la confiablidad de los métodos de relleno: U.S National Weather Servicie (WS), deductivo racional (RD), la regresión múltiple (RM) y regresión lineal (RL) y de ellos utilizar el mejor para rellenar los datos de las series de precipitación, temperatura máxima y temperatura mínima en el periodo 2006-2009 de las estaciones ubicadas en la zona Norte del eje bananero del Urabá Antioqueño. La estación base fue la estación Pista Indira, y las estaciones Uniban, Turbo, Aeropuerto y Prado Mar se utilizaron como vecinas. La estimación del error se hizo con la: raíz cuadrada del cuadrado medio del error (RCCME), coeficiente de determinación (R 2 ), error medio absoluto (MAE), error relativo (RE) e Índice de concordancia de Willmott (d). Los métodos RM y RL presentaron RCCME y MAE similares a los del WS, lo que con llevaría a tener errores similares pero debido a que la R 2 de los primeros fueron menores de 0.8 y el índice (d) para WS fue mayor o similar a lo demás método, se utilizó este último para el relleno de las series.Palabras claves: deductivo, distancia inversa, regresión. AbstractAgroclimatic studies require the use of time series of meteorological variables, which generally present missing data limiting its use, so filling methods are used. The purpose of this study was to determine the reliability of filling methods: US National Weather Servicie (WS), rational deductive (RD), multiple regression (MR) and linear regression (LR) and from them use the best to fill the data series of precipitation, maximum and minimum temperature during the period 2006-2009 of the stations located in the northern area on b anana axis from Urabá Antoquia. The base station was Pïsta Indira, and Uniban, Turbo, Aeropuerto and Prado Mar stations were used as neighbors. Standard error was made with: square root of the mean squared error (RCCME), coefficient of determination (R 2 ), mean absolute error (MAE), relative error (RE) and Willmott's concordance index (d). RM and RL methods presented similar RCCME and MAE to those of WS, which would lead to have similar errors but due to R 2 of the first were 0.8 and the index (d) for WS was higher or similar to the other method, the latter was used to fill in the series.Keywords: deductive, inverse distance, regression. IntroducciónEl clima es un recurso natural que afecta la producción agraria, por lo tanto es importante tener conocimiento de los recursos disponibles del ambiente. Las interacciones que se producen en el área del cultivo, debajo de la superficie del suelo, la interfase suelo-aire y la capa límite de la atmósfera, proporciona una guía esencial para las decisiones estratégicas en planificación a largo plazo de los sistemas agrícolas (WMO, 1981).En las estaciones climatológicas es común que los registros estén inco...

show abstract

Comparison of missing value imputation methods in time series: the case of Turkish meteorological data

Cited by 146 publications

References 52 publications

Dealing with missing data for prognostic purposes

Dealing with missing data for prognostic purposes

A Two-Step Method for Missing Spatio-Temporal Data Reconstruction

Relleno de series diarias de precipitación, temperatura mínima, máxima de la región norte del Urabá Antioqueño

Contact Info

Product

Resources

About