Tutorial 1

Tutorial 1  Machine Learning for Data-Driven Smart Grid Applications &
 Detection of Irregular Power Usage using Machine Learning
Presenters Patrick GLAUNER, Jorge Augusto MEIRA and Radu STATE, University of Luxembourg
Date and Time 22 May 2018, 08:00 am – 12:00 pm
Location  MR308
Details

i. Machine Learning for Data-Driven Smart Grid Applications

Abstract
The field of Machine Learning grew out of the quest for artificial intelligence. It gives computers the ability to learn statistical patterns from data without being explicitly programmed. These patterns can then be applied to new data in order to make predictions. Machine Learning also allows to automatically adapt to changes in the data without amending the underlying model. We deal every day dozens of times with Machine Learning applications such as when doing a Google search, using spam filters, face detection, speaking to voice recognition software or when sitting in a self-driving car. In recent years, machine learning methods have evolved in the smart grid community. This change towards analyzing data rather than modeling specific problems has led to adaptable, more generic methods, that require less expert knowledge and that are easier to deploy in a number of use cases. This is an introductory level course to discuss what machine learning is and how to apply it to data-driven smart grid applications. Practical case studies on real data sets, such as load forecasting, detection of irregular power usage and visualization of customer data, will be included. Therefore, attendees will not only understand, but rather experience, how to apply machine learning methods to smart grid data.

Contents

  • Introduction to machine learning: Supervised, unsupervised and reinforcement methods, History, Delimitation to related fields such as data mining, data science, etc.

  • Supervised learning: Regression: linear regression, higher-order regression models, Classification: decision trees, logistic regression, support vector machines, neural networks, Model selection.

  • Unsupervised learning: Clustering, Dimensionality reduction.

  • Recent advances in machine learning: Deep Learning, General comparability of machine learning approaches (no free lunch theorem).

  • Practical case studies on real data sets: Visualization of customer data, Load forecasting, Estimation of power losses.

Intended Audience
This tutorial assumes no prior experience in machine learning or analysis of smart grid data. Attendees will be able to acquire both the theoretical foundations and an understanding of real- world data-driven smart grid applications based on machine learning in this tutorial. Attendees with prior experience in machine learning or smart grid data will benefit from this tutorial by experiencing a comprehensive rehearsal of the theoretical foundations as well as new real-world insights. These attendees will also be able to contribute to discussions by sharing their prior experience.

Special Requirements
Attendees should bring a laptop for live demos. We will use Python-a high-level programming language, and the scikit-learn machine learning package. We will provide instructions, such that attendees can install both in advance.

ii. Detection of Irregular Power Usage using Machine Learning
Abstract
Electricity losses are a frequently appearing problem in power grids. Non-technical losses (NTL) appear during distribution and include, but are not limited to, the following causes: Meter tampering in order to record lower consumptions, bypassing meters by rigging lines from the power source, arranged false meter readings by bribing meter readers, faulty or broken meters, un-metered supply, technical and human errors in meter readings, data processing and billing. NTLs are also reported to range up to 40% of the total electricity distributed in countries such as India, Pakistan, Malaysia, Brazil or Lebanon. This is an introductory level course to discuss how to predict if a customer causes a NTL. In the last years, employing data analytics methods such as machine learning and data mining have evolved as the primary direction to solve this problem. This course will present and compare different approaches reported in the literature. Practical case studies on real data sets will be included. As an additional outcome, attendees will understand the open challenges of NTL detection and learn how these challenges could be solved in the coming years.

Contents

  • Introduction to NTL: Definition of NTL, Impact on economies and grids.

  • Overview about machine learning and data mining.

  • State-of-the-art: Comprehensive comparison of research works presented in the literature based on machine learning, Contrast to other approaches such as expert systems, energy balance, etc., Comparison of data sets and evaluation metrics used in NTL research.

  • Practical case studies on real data sets: Analyzing consumption data for predicting NTL, Handling biases in the data, such as inspection results, combining machine learning with expert knowledge in order to reduce inspection costs.

  • Discussion of open challenges to solve in order to advance NTL detection.

Speakers Bio

Patrick GLAUNER is a PhD student at the University of Luxembourg working on the detection of electricity theft through machine learning. He graduated as valedictorian from Karlsruhe University of Applied Sciences with a BSc in computer science and obtained his MSc in machine learning from Imperial College London. He was a CERN Fellow, worked at SAP and is an alumnus of the German National Academic Foundation (Studienstiftung des deutschen Volkes). He is also adjunct lecturer of artificial intelligence at Karlsruhe University of Applied Sciences. His current interests include anomaly detection, big data, computer vision, deep learning and time series.

Jorge Augusto MEIRA was born in Brazil in 1983. He received the Ph.D. degree in computer science from the University of Luxembourg, Luxembourg, in 2014. He has held an Assistant Professor position at Federal University of Paran, Brazil, in 2015. His main areas of research interest are Software Testing, Databases Systems and Big Data. He is currently a Research Associate at the University of Luxembourg working on detection of non-technical losses.

Radu STATE heads the research group SEDAN (Service and Data Management in Distributed Systems) in the Interdisciplinary Centre for Security, Reliability and Trust, University of Luxembourg. He holds a Master of Science degree from the Johns Hopkins University, USA and a PhD degree obtained during his research activity with INRIA, France. He was a Senior Researcher at INRIA, France and Professor of Computer Science at Telecom Nancy, France.