A Novel Missing Data Imputation Approach for Time Series Air Quality Data Based on Logistic Regression

Chen, Mei; Zhu, Hongyu; Chen, Yongxu; Wang, Youshuai

doi:10.3390/atmos13071044

Cited by 13 publications

(3 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Detecting the missingness mechanism is considered an important step for manipulating MVs. This paper considers and deals with the three kinds of missingness mechanisms [5][6][7][8].…”

Section: Missingness Mechanismsmentioning

confidence: 99%

Handling Missing Values Based on Similarity Classifiers and Fuzzy Entropy Measures

et al. 2022

View full text Add to dashboard Cite

Handling missing values (MVs) and feature selection (FS) are vital preprocessing tasks for many pattern recognition, data mining, and machine learning (ML) applications, involving classification and regression problems. The existence of MVs in data badly affects making decisions. Hence, MVs have to be taken into consideration during preprocessing tasks as a critical problem. To this end, the authors proposed a new algorithm for manipulating MVs using FS. Bayesian ridge regression (BRR) is the most beneficial type of Bayesian regression. BRR estimates a probabilistic model of the regression problem. The proposed algorithm is dubbed as cumulative Bayesian ridge with similarity and Luca’s fuzzy entropy measure (CBRSL). CBRSL reveals how the fuzzy entropy FS used for selecting the candidate feature holding MVs aids in the prediction of the MVs within the selected feature using the Bayesian Ridge technique. CBRSL can be utilized to manipulate MVs within other features in a cumulative order; the filled features are incorporated within the BRR equation in order to predict the MVs for the next selected incomplete feature. An experimental analysis was conducted on four datasets holding MVs generated from three missingness mechanisms to compare CBRSL with state-of-the-art practical imputation methods. The performance was measured in terms of R2 score (determination coefficient), RMSE (root mean square error), and MAE (mean absolute error). Experimental results indicate that the accuracy and execution times differ depending on the amount of MVs, the dataset’s size, and the mechanism type of missingness. In addition, the results show that CBRSL can manipulate MVs generated from any missingness mechanism with a competitive accuracy against the compared methods.

show abstract

“…Detecting the missingness mechanism is considered an important step for manipulating MVs. This paper considers and deals with the three kinds of missingness mechanisms [5][6][7][8].…”

Section: Missingness Mechanismsmentioning

confidence: 99%

Handling Missing Values Based on Similarity Classifiers and Fuzzy Entropy Measures

et al. 2022

View full text Add to dashboard Cite

show abstract

“…In this context the [12] categorizes the mechanisms to handle incomplete data in two methods: statistical and machine learning (ML). The statistical approaches are the oldest techniques utilized for estimation such as regression model [13], Hidden Markov Model (HMM) [14] and Expectation Maximization (EM) [15]. The ML approach is categorized into two categories: supervised and unsupervised.…”

Section: Introductionmentioning

confidence: 99%

A New Imputation Technique Based a Multi-Spike Neural Network to Handle Missing Data in the Internet of Things Network (IoT)

Al-Jamali,

Al-Saedi,

Zarzoor

et al. 2023

IEEE Access

View full text Add to dashboard Cite

Over the past decade, the Internet of Thing (IoT) devices have been deployed in wide-scale several applications to collect vast amount of data from different locations in a time-series manner. However, collected data may be missing or damaged due to several issues such as unreliable communications, faulty sensors and synchronization problem that decrees application accuracy. Therefore, a several imputation-based machine learning approaches have been suggested to handle this problem in IoT application. In this study, a new approach is proposed called impute missing data (IMD) based on multi-Spike Neural Network learning method called IMD-SNN, to increase the reliability of missing value imputation in IoT. The method consists of three phases: Inserting missing data, to evaluate the missing values based on the cumulative distribution function (CDF), the multi SNN phase to estimate missing data according to the timestamp and a performance evaluation phase to evaluate an imputation accuracy via made a comparison with two models: imputation based KNN (I-KNN) and Imputation based (I-MLP) model based on resource usage and imputation accuracy assessment metrics. The implementation results have been shown that IMD-SNN utilizes less energy usage in comparison with (I-MLP) model and I-KNN model and gives highest imputation accuracy in contrast with (I-MLP) model and I-KNN model. Also, the IMD-SNN model utilizes less memory usage and needs execution time less than I-MLP model. INDEX TERMS Internet of Thing (IoT), Spike Neural Network (SNN), Multilayer Perceptron (MLP), Root Mean Square Error (RMSE)

show abstract

“…Exposure to higher concentrations of fine particulate matter (PM2.5) is a major public health concern across the world. This is mainly because of their small size, PM2.5 can penetrate deep into the lungs, heart, and bloodstream causing detrimental health effects (Chen et al, 2022). The health effects of PM2.5 range from respiratory diseases, cardiovascular diseases, and cancer to mortality in worst cases (Wang et al, 2021) (Ahani et al, 2020).…”

Section: Introductionmentioning

confidence: 99%

Temporal Characteristics and Health Risks Assessment of PM2.5 in Ambient Air in Thohoyandou, South Africa

Novela,

Gitari,

Mudzielwana

et al. 2023

Preprint

View full text Add to dashboard Cite

This study determined the temporal variation of PM2.5 in ambient air in Thohoyandou and further assessed the associated health risks. The levels of PM2.5 were quantified for a period of 1 year (April 2017-April 2018) using the gravimetric method. There was no significant difference (P-value = 0.18) in concentrations of both PM2.5 samples collected during weekdays (11.29 µg.m-3) and weekends (9.86 µg.m-3). However, higher concentrations of PM2.5 were measured in spring and the lowest was measured in summer. The cancer risk obtained for PM2.5 (2.21 × 10−5, 3 × 10-4, and 5 × 10-4 for infants, children, and adults respectively) in the outdoor air of Thohoyandou has exceeded the limit values by the USEPA and WHO, implying a significant risk for the whole population. For non-carcinogenic risks, the HQ values were 2.60, 4.81, and 2.60 for infants, children, and adults respectively. The HQ value &gt;1 indicates a non-carcinogenic risk to the residents in Thohoyandou and a higher risk to children. Moreover, PM2.5 in Thohoyandou is responsible for 0.15% and 0.13% of deaths resulting from cardiovascular disease and lung cancer respectively for adults above 30 years. PM2.5 is causing adverse health effects in Thohoyandou as deduced from the health risk assessment. Therefore, it is recommended that further epidemiological studies be conducted in Thohoyandou to estimate the burden of disease due to exposure to particulate matter and suitable controlling policies be arranged to reduce particulate matter.

show abstract

A Novel Missing Data Imputation Approach for Time Series Air Quality Data Based on Logistic Regression

Cited by 13 publications

References 44 publications

Handling Missing Values Based on Similarity Classifiers and Fuzzy Entropy Measures

Handling Missing Values Based on Similarity Classifiers and Fuzzy Entropy Measures

A New Imputation Technique Based a Multi-Spike Neural Network to Handle Missing Data in the Internet of Things Network (IoT)

Temporal Characteristics and Health Risks Assessment of PM2.5 in Ambient Air in Thohoyandou, South Africa

Contact Info

Product

Resources

About