Outlier Detection in Multivariate Time Series Data Using a Fusion of K-Medoid, Standardized Euclidean Distance and Z-Score

Chikodili, Nwodo Benita; Mohammed, Abdulmalik Danlami; Abisoye, Opeyemi Aderiike; Bashir, Sulaimon Adebayo

doi:10.1007/978-3-030-69143-1_21

Cited by 17 publications

(6 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The z-scores were computed for all study variables to detect potential outliers. The z-scores larger than +3 or lower than −3 would be considered outliers and excluded from the analysis (Chikodili et al, 2021).…”

Section: Discussionmentioning

confidence: 99%

Does sports participation affect the mediating role of impulsivity in the association between adverse childhood experiences and aggression?

Janković,

Van Boxtel,

Bogaerts

2024

Front. Psychol.

View full text Add to dashboard Cite

IntroductionAdverse childhood experiences (ACEs) and self-reported aggression have often been associated, however, the mediating and moderating mechanisms underlying this association are not fully understood. In addition, less is known about potential factors that could serve as protection against the development of aggressive behavior. In this study, we investigated a moderated mediation model of sports participation and five impulsivity traits including positive and negative urgency, sensation seeking, and lack of premeditation and perseverance, in the association between ACEs and two forms of aggression.Methods and resultsThe sample comprised 651 (68.5% females) individuals from a Dutch community ranging in age from 18 to 82 years (M = 34.08). Contrary to our expectations, sports participation did not affect the associations of ACEs, the five impulsivity traits, and reactive and proactive aggression, except the association between ACEs and lack of premeditation. Specifically, the detrimental effect of ACEs on the propensity to act without thinking, increases among individuals with lower levels of sports participation. In addition, it was also found that sports participation directly contributed to higher levels of sensation seeking. Finally, after controlling for sociodemographic variables, the positive association between ACEs and reactive aggression was significantly mediated by negative urgency, lack of perseverance, and lack of premeditation, while the positive association between ACEs and proactive aggression was significantly mediated by lack of premeditation and positive urgency.ConclusionThe findings of this study add to the body of knowledge about the role of sports participation and impulsivity traits in the development of both forms of aggression, however, replication studies among multiple populations are needed before firm conclusions can be drawn.

show abstract

Section: Discussionmentioning

confidence: 99%

Does sports participation affect the mediating role of impulsivity in the association between adverse childhood experiences and aggression?

Janković,

Van Boxtel,

Bogaerts

2024

Front. Psychol.

View full text Add to dashboard Cite

show abstract

“…This involved handling missing data through techniques like imputation and deletion (considering potential biases from exclusion). We employed robust outlier detection and removal methods (z-scores, winsorization) [72] from Pandas and SciPy libraries to mitigate their influence. Finally, the study performed data validation to ensure consistency and alignment with the study's objectives, including cross-referencing [73] with external sources where applicable.…”

Section: Vhi Methodsmentioning

confidence: 99%

Resilience and Adaptation Strategies in Response to Drought: A Spatial and Community-Based Study in West Pokot, Kenya

Munyaka,

Gallay,

Abdem

et al. 2024

Preprint

View full text Add to dashboard Cite

This research provides an in-depth analysis of drought impacts in West Pokot, Kenya, utilizing both remote sensing surveys and focus group discussions. The study commenced by assessing the impact of drought on vegetation health via spatial sensors. It then utilized surveys and focus group discussions (FGDs) to engage farmers from the most drought-affected areas in Kacheliba and Sigor sub-counties. The analysis of remote sensing data from 1990 to 2022 highlighted that Sigor and Kacheliba experienced significant drought events during the years 1995, 2000, 2002, 2005, and 2009, and 1991, 1995, 2000, 2005, and 2019, respectively. This study delves into both the meteorological and socio-economic effects on local communities. Using the Multinomial Logistic Regression (MLR) model with the glm function, the research uncovered the impacts of so-cio-economic factors and agricultural practices on household livestock production. Additionally, the study employed Multiple Correspondence Analysis (MCA) to explore the dynamics between livestock production, community resilience, and crop production. The findings underscore the urgent need for tailored interventions to enhance community resilience in the face of increasing climatic extremes, serving as a blueprint for similar arid and semi-arid regions worldwide.

show abstract

“…Figure 2(a) shows the boxplot marking the median, quartiles, IQR, and outliers. The application of Z-score as an outlier detection technique is common in different sectors; a few examples include the work of [12], [13]. These researchers have successfully applied the Z-score algorithm in high-dimensional and time-series data.…”

Section: Outlier Identi Cation Using ML Methodsmentioning

confidence: 99%

Application of Machine-Learning Algorithms in Outlier Detection in Well Logs

Talreja

Khan

2022

Preprint

View full text Add to dashboard Cite

The digital transformation in the oil and gas sector began because of two economic downturns. Literature review suggests an increasing work on outlier identification in geophysical well logs with some Exploration and Production companies have started working internally on this aspect as it consumes a significant amount of time for a geoscientist. While performing 1D geomechanical modelling, data preparation is the most time-consuming step, often demanding close to 4 to 5 hours. The data preparation steps include removing the outliers present in the well log data caused by the casing shoe, bad hole intervals, tool noise, and many other factors. Geoscientists invest hours to clean these outliers and handle the resulting missing values. Identification of outliers is mostly manual and missing data are replaced with simple methods based on linear, power, or exponential algorithms derived in a laboratory setup. In this work, seven different anomaly detection techniques were evaluated on well logs. The proposed workflow includes a combination of Z-score and Isolation Forest algorithms to clean and condition the data and handle missing values. The workflow then constructed the cleaned composite log data in quick time, which was then directly consumed by 1D geomechanical workflows. This workflow is a handy tool that adds value by saving time in data preparation. This study will redound to the benefit of geoscientists by saving significant amount of the time spent in the identification of outliers and hence reducing the overall turnaround time of modelling.

show abstract

Outlier Detection in Multivariate Time Series Data Using a Fusion of K-Medoid, Standardized Euclidean Distance and Z-Score

Cited by 17 publications

References 15 publications

Does sports participation affect the mediating role of impulsivity in the association between adverse childhood experiences and aggression?

Does sports participation affect the mediating role of impulsivity in the association between adverse childhood experiences and aggression?

Resilience and Adaptation Strategies in Response to Drought: A Spatial and Community-Based Study in West Pokot, Kenya

Application of Machine-Learning Algorithms in Outlier Detection in Well Logs

Contact Info

Product

Resources

About