Kernel Function Selection for FISVDD Applied to Outliers Detection Application for Energy Time Series Datasets

Abstract

Forecasting energy consumption is crucial for maintaining stability in the energy supply-demand balance. This process requires accurate data, necessitating deep preprocessing steps, with outlier detection being a pivotal task. While most existing methods are tailored for batch learning, smart grid data collection operates continuously, demanding online algorithms for effective data cleaning. This paper proposes an analytical study emphasizing the significance of selecting an appropriate kernel function for data description methods and its impact on forecasting efficiency. The Fast Incremental Support Vector Data Description (FISVDD) algorithm is chosen as a kernel-based approach for outlier detection in time series datasets. The evaluation involves different kernel functions on two datasets: Electrical Grid Stability Simulated Data and Individual Household Electric Power Consumption Dataset. Criteria such as the objective function and data distribution are considered to determine the most suitable kernel function. FISVDD's performance with the right kernel function is compared with other outlier detection methods, and the results are fed into a Gated Recurrent Unit model to visualize the impact on forecasting. Mean Squared Error evaluation reveals that FISVDD with the appropriate kernel function yields the best results, underscoring the strength of data description methods with the appropriate kernel function.

Keywords:

• outliers detection
• incremental learning
• FISVDD
• kernel functions
• kernel based methods
• energy time series datasets

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Ibtissam Amalou

Ibtissam Amalou is a doctoral student at the Faculty of Science and Technology of Cadi Ayyad University, focusing on the prediction of energy consumption through artificial intelligence approaches. She holds a Master's degree in Data Science and Decision Support from the same university. With a strong background in computer science, Ibtissam also serves as a computer science instructor at the Moroccan School of Engineering Sciences. Her academic pursuits and professional roles reflect her dedication to advancing knowledge in the field of energy consumption prediction and artificial intelligence.

Naoual Mouhni

Naoual Mouhni is currently a Professor at University Moulay Ismail, where she brings a wealth of experience and expertise to the academic realm. Formerly, she held the role of Director of Research and Development at EMSI Marrakech and spearheaded operations at the LAMIGEP multidisciplinary research lab. Prior to her professorship, she served as a Professor of Computer Science at the same institution since 2012.Research-wise, Naoual MOUHNI is deeply engaged in advancing the frontiers of artificial intelligence, renewable energies, and database technologies. Her commitment to pushing these fields forward is evident through her ongoing contributions and exploration within each domain.

Abdelmounaim Abdali

Abdelmounaim Abdali PhD in Solid Mechanics and Structures in University Picardie Jules Vernes of Amiens in 1996, France. He is a Professor of Computer Science, Deputy Director of the Control and Computer Science for Intelligent Systems and Green Energy (CISIEV) team at the Faculty of Science and Technology in Marrakech, Cadi Ayyad University (Morocco). My research interests include Big Data, data science, energy optimization in smart grids, DTN mobile networks, bone remodeling, data analysis, Machine Learning, Deep Learning and Numerical simulations.