=Paper=
{{Paper
|id=Vol-2874/short12
|storemode=property
|title=A Review on Latest Trends in Non-Technical Loss Detection
|pdfUrl=https://ceur-ws.org/Vol-2874/short12.pdf
|volume=Vol-2874
|authors=Khawaja MoyeezUllah Ghori,Muhammad Awais,Akmal Saeed Khattak,Muhammad Imran,Rabeeh Ayaz Abbasi,Laszlo Szathmary
}}
==A Review on Latest Trends in Non-Technical Loss Detection==
https://ceur-ws.org/Vol-2874/short12.pdf
A Review on Latest Trends in
Non-Technical Loss Detection∗
Khawaja MoyeezUllah Ghoriab , Muhammad Awaisc ,
Akmal Saeed Khattakd , Muhammad Imrane ,
Rabeeh Ayaz Abbasif , Laszlo Szathmaryg
a
Department of Computer Science, National University of Modern Languages, NUML,
Islamabad, Pakistan
mghouri@numl.edu.pk
b
University of Debrecen, Doctoral School of Informatics, Debrecen, Hungary
c
Department of Computer Science, Edge University, Ormskirk, United Kingdom
mawais@ieee.org
d
Department of Computer Sciences, Quaid-i-Azam University, Islamabad, Pakistan
akhattak@qau.edu.pk
e
College of Applied Computer Science, King Saud University, Riyadh, Saudi Arabia
dr.m.imran@ieee.org
f
Department of Computer Sciences, Quaid-i-Azam University, Islamabad, Pakistan
rabbasi@qau.edu.pk
g
University of Debrecen, Faculty of Informatics, Debrecen, Hungary
szathmary.laszlo@inf.unideb.hu
Proceedings of the 1st Conference on Information Technology and Data Science
Debrecen, Hungary, November 6–8, 2020
published at http://ceur-ws.org
Abstract
An increasing interest in digging out the consumption patterns in power
and energy sector is observed globally. This includes electrical, gas, and wa-
ter supply industries. A reason behind analyzing the consumption patterns
Copyright © 2021 for this paper by its authors. Use permitted under Creative Commons License
Attribution 4.0 International (CC BY 4.0).
∗ The work/publication is supported by the EFOP-3.6.1-16-2016-00022 project. The project is
co-financed by the European Union and the European Social Fund.
131
is the detection of fraudulent attempts which are made for the illegal reduc-
tion of bill payments. In the case of electricity, these attempts are made by
reversing the meters, by-passing or slowing down the meters or inaccurate
readings. The detection of theft attempts in power industry is termed as
Non-Technical Loss (NTL) detection. With the increasing demand for elec-
tricity, the occurrences of NTL have been reported globally including India,
Pakistan, Brazil and China etc. In this paper, we first describe the use of the
synthesized and the real datasets in NTL detection. Then, we highlight an
interesting characteristic of class imbalance that is exhibited in the datasets
used for NTL detection. Moreover, we identify the fruitful areas in NTL
detection where the research community has been working on. Lastly, we
discuss the need for a relative comparison of the classical machine learning
and deep learning over a benchmark dataset for NTL detection.
Keywords: Non-Technical Loss (NTL), Non-Technical Loss detection, ma-
chine learning, classification, class imbalance
1. Introduction
Recently, an increasing interest has been observed in recognizing the consumption
patterns of the consumers of electricity, gas and water supplies [25]. One of the main
objectives of this activity is to identify and forecast the potential theft attempts
in order to have reduced bills. This illegal theft attempt has dented the economies
of many countries causing a loss of billions of dollars. This includes China [14],
Pakistan [10], India, Brazil [18], etc.
Non-Technical Loss (NTL) detection in electric power industry is a term used
for the detection of faulty meters or illegal usage of electric units. Losses are bore
by the electric supply companies on the account of faulty meters that record fewer
units as compared to the consumed electricity. On the other hand, this practice
can also be intentional in order to get the electricity bill reduced by a substantial
margin. For both cases, the supplier companies look for a solution which can
identify them the faulty meters or potential theft instances.
One of the important characteristics of the relevant datasets is that they be-
long to the class imbalance problem. It is the problem where the dataset is biased
towards one class by its heavy representation while the other class is less represen-
tative. Interestingly, the problem becomes more challenging when the focus is on
the true representation of the less representative class. Naturally, the number of
normal electric consumption in a neighborhood area is huge as compared to the
number of theft attempts. This gives a clear indication that the real dataset of
the consumption of electricity belongs to the class imbalance problem where the
number of negative class samples is huge as compared to the number of positive
class samples. The techniques used in NTL detection should be able to balance
out the positive and the negative class samples before the dataset is used by the
machine learning algorithms [9].
One of the techniques used to identify the NTL is applying classical machine
learning algorithms to the datasets pertaining to the consumption of electricity.
132
This includes the use of Support Vector Machine (SVM), KNN, decision trees,
ensemble methods and neural networks [8]. Advances in deep learning has attracted
some researchers to test different variants of deep learning for NTL detection. For
e.g., the authors of [3] have used deep neural networks along with long short-term
memory network to identify the occurrences of NTL in a dataset pertaining to
the smart meters of a utility company in Spain. However, there is still a need
to compare the performances of the classical machine learning algorithms with
the different variants of deep learning architecture. In this paper, we focus on
elaborating the importance of a comparative study of the two paradigms for NTL
detection in a real dataset.
2. Synthesized Vs. Real Datasets
Two types of datasets are used in NTL detection. One of the types belongs to the
synthesized datasets which are randomly generated keeping in mind the require-
ments of NTL detection. One of the benefits of using the synthesized datasets
is that they are easily accessible while on the other hand, they might miss the
potentially useful information which otherwise would have been handy in detect-
ing unlawful consumption activities. The other type is the real dataset which is
taken from a distribution company. An essential advantage of using such datasets
is analyzing the real and unique patterns of the consumption of electricity which
might be missing in the synthesized datasets. An associated drawback of using
real datasets is that it is hard to get an access of such datasets as distribution
companies generally avoid sharing the consumer’s information.
3. Class Imbalance: An aspect of NTL Detection
An interesting property associated with the datasets of electric distribution com-
panies is that they belong to the class imbalance problem. Considering two classes;
normal consumption and abnormal consumption; the class imbalance problem
arises when most of the consumption records belong to the normal consumption
class while few records belong to the abnormal consumption class. When the ob-
jective is predicting the abnormal consumption class which is rarely represented
in the dataset, the success ratio of the classifiers might get deteriorated unless the
class imbalance ratio is properly dealt.
To balance out the number of representative records for both the classes, under-
sampling the majority class can be used in the pre-processing stage of data prepara-
tion. This is termed as synthetic minority over-sampling technique (SMOTE) [30].
The other way is over-sampling the minority class, i.e., duplicating or randomly
creating new synthesized records of the minority samples [4]. Both techniques have
been used in different problem domains belonging to the class imbalance. However,
few have tried these techniques in NTL detection.
133
4. Which Areas in NTL Detection are fruitful?
There is a general pattern followed in the pre-processing of the datasets and appli-
cation of the classifiers in order to achieve a significant success in NTL detection.
This pattern is shown in the Figure 1. The electricity distribution companies pro-
vide raw data containing consumption records. Depending on the type of metering
infrastructure, the consumption data is either half hourly, hourly, daily, bi-monthly
or monthly records. The automated metering infrastructure (AMI) facilitates to
record hourly or daily consumption while the manual metering infrastructure uses
monthly manual readings done by the meter readers. Not all the records in con-
sumption profile are useful in performing analytics for NTL detection. Similarly,
not all the features are important for detecting NTL. For this, record selection and
feature selection is performed. Once the dataset is ready, it is scaled for normaliz-
ing the values. The next step is training and testing of the classifiers. A range of
classifiers of different types is applied. The performance of the classifiers is then
measured by performance evaluation metrics.
Synthesized Vs.
Real Dataset
Best Features
Monthly
Raw Data Data Munging/
Consumption Feature Selection
Record Selection
Records
Post Processing Performance Training Scaling for
and NTL Evaluation and Testing Feature
Detection Metrics Classifiers Normalization
Best Metrics
for NTL Best
Detection Classifiers
Figure 1. Pre-processing Pattern of NTL Detection.
The researchers in this area are largely interested in three main modules. Some
of them try to find the best combination of features which are most suited for
NTL detection. For this, they use feature importance. In one of our previous
contributions, we propose the Incremental Feature Selection (IFS) algorithm that
selects the best combination of features responsible in detecting the occurrence of
NTL [9]. Some authors have taken a keen interest in finding the best individual
classifiers or a combination of classifiers for NTL detection. Others have tried to
find the best type of the classifiers for NTL detection.
Some authors have focused on the performance evaluation module of the clas-
sifiers. They have tried to find out which metrics are best suited to evaluate the
classifiers considering the specifications of the NTL detection problem. For exam-
134
ple, we have concluded in one of our previous works that recall should be given a
higher priority as compared to any other performance evaluation metric considering
the special characteristics of the NTL detection problem [10].
5. Machine Learning Vs. Deep Learning in NTL De-
tection
Machine learning is mainly used for Non-Technical Loss (NTL) detection. NTL is
a typical classification task. Most of the researchers have used unsupervised, super-
vised, semi-supervised, hybrid and network based methods. Some of the research
on NTL detection can be seen in [2, 11–13, 15, 17, 19, 23, 24, 26–29, 31–36]. Table 1
summarizes research work on NTL detection. Most researchers used electric supply
and distribution companies of specific countries as it is shown in Table 1. The most
common evaluation measures used in NTL detection are accuracy, precision, recall,
AUC and others as Table 1 shows. A framework to detect NTL is proposed in [17].
The characteristics of consumers were analyzed and data mining techniques were
used to recover electrical energy loss. Consumption data, extracted from a Spanish
power supply firm known as Endesa Distribution, was used by [17]. Association
rule mining was used to find groups of consumers responsible for non-technical loss
in the form of electricity theft. Hartmann et. al. [13] used contextual learning on a
dataset from Luxembourg electric distribution company. Accuracy, precision, recall
and F1 measures were used as evaluation measures. The attributes of consumption
were investigated in [26] for NTL detection in two firms. For this purpose, hierar-
chical clustering, Benford curve and Multi-Dimensional Scaling (MDS) were used
on two datasets based on two Colombian companies. The study was evaluated using
ROC. Local Outlier Factor (LOF) based on the DBSCAN clustering algorithm was
used in the research work of [28]. The technique was evaluated using the Silhouette
coefficient and Davies Bouldin index. The research work of [34] used a synthetic
dataset and observed unusual profiles of electricity consumers by proposing a dis-
tance matrix. Area Under ROC Curve (AUC), F-1 measure and accuracy was
used as evaluation measure. The results showed the effectiveness of the proposed
technique after comparing it with DBSCAN, GMM and kNN. Yeckle and Tang
[33] carried experiments on Irish dataset. Different outlier detection methods are
used to detect NTL. The performance is measured by AUC. The research work of
Zheng et. al. [36] explored deep convolutional neural networks (CNN) on a dataset
based Chinese electricity company. The result from deep convolutional neural net-
works is compared with SVM, random forest, logistic regression and TSR (Three
Sigma Rule). It was observed that deep CNN outperformed the above mentioned
classifiers. Another research work of [32] used 4000 Irish household data records
collected from smart metering. Gustafson-Kessel clustering algorithm was used in
[32] to discriminate non-technical loss. The rate of true positive was observed as
63.6% whereas false positive was 24.3%.
A dataset collected from a Spanish electrical supply firm was analyzed in the
135
Table 1. Literature review on NTL detection.
Year Author-Ref. Data Techniques Evaluation
2015 Hartmann et. al. - [13] Luxembourg contextual Accuracy,
electric dis- learning, precision,
tribution live machine recall, f1
company learning, measure
Gaussian
mixture
model, pro-
file power
consumption
2016 Peng et. al. - [24] Chinese elec- Fuzzy clus- Mean Index
tric company tering and Adequacy
classification and execu-
tion time
(MIA)
2017 Sánchez et. al - [26] Two Colom- Hierarchical ROC
bian Compa- Clustering,
nies Multidi-
mensional
Scaling, De-
cision Trees
2017 Sharma et. al - [28] Data collec- Local Out- Silhouette
tion from lier Factor coefficient,
USA and (LOF), Davies
India (DBSCAN) Bouldin in-
clustering dex
algorithm
2017 Zheng et. al - [34] Synthetic density- Area Under
dataset based elec- ROC Curve
tricity theft (AUC), ac-
detection curacy and
F1 measures
2018 J. Yeckle and B. Tang - [33] Irish dataset Outlier AUC
detection
techniques
2018 Zheng et. al. - [36] Chinese Deep CNN AUC
electricity
dataset
2018 Viegas et. al. - [32] Irish house- Gustafson- True positive
hold data Kessel fuzzy and false
collected clustering positive rate
from smart algorithm
metering
2018 Guerrero et. al. - [11] Spanish ANN, classi- Accuracy
electricity fication, re-
company gression and
data SOP
2020 Pazi et. al - [23] Electrical SVM, kNN Accuracy,
consumption and naïve detection
data for a Bayes rate, preci-
large mu- sion, true
nicipality in negative rate
South Africa
136
research of [11]. Artificial neural network, regression and Self Organizing Map
(SOM) are used and as a result it was claimed that the accuracy has improved.
Statistical learning methods were used in [23] to detect NTL in South Africa. Three
classifiers namely Support Vector Machine (SVM), kNN (k Nearest Neighbour) and
naïve Bayes were used. The classification models were evaluated using accuracy,
detection rate, precision and true negative rate.
Less attention has been paid to deep neural networks and network science based
algorithms to detect NTL. From the literature review presented, it can be observed
that classical machine learning is the major streamline in solving the problem
of NTL detection. However, there is still a need of a thorough testing of different
variants of deep learning in a real dataset for NTL detection. Community detection
techniques have a key role in computer science, sociology, biology, physics,
economics, engineering, marketing, ecology, political sciences and many other
fields [7]. Community detection is used for inferring useful information from
complex networks. Complex networks are getting attention from researchers of
different domains. A network such as biological, information, technological,
social and other can be modelled as graphs [6]. Some well known community
detection techniques such as Greedy Modularity Maximization [5, 16, 20], Girvan-
Newman algorithm [21], Louvain algorithm [1] and k-clique percolation [22] can also
be used to identify interesting patterns in NTL.
References
[1] V. D. Blondel, J.-L. Guillaume, R. Lambiotte, E. Lefebvre: Fast unfolding of com-
munities in large networks, Journal of Statistical Mechanics: Theory and Experiment 2008.10
(Oct. 2008), P10008,
doi: 10.1088/1742-5468/2008/10/p10008.
[2] M. Buevich, X. Zhang, O. Shih, D. Schnitzer, T. Escalada, A. Jacquiau-Chamski,
J. Thacker, A. Rowe: Microgrid Losses: When the whole is Greater than the Sum of its
parts, in: Proceedings of the 7th International Conference on Cyber-Physical Systems, IEEE
Press, 2016, pp. 46–50.
[3] M.-M. Buzau, J. Tejedor-Aguilera, P. Cruz-Romero, A. Gomez-Exposito: Hybrid
deep neural networks for detection of non-technical losses in electricity smart meters, IEEE
Transactions on Power Systems 35.2 (2019), pp. 1254–1263.
[4] N. V. Chawla, K. W. Bowyer, L. O. Hall, W. P. Kegelmeyer: SMOTE: synthetic mi-
nority over-sampling technique, Journal of artificial intelligence research 16 (2002), pp. 321–
357.
[5] A. Clauset, M. E. J. Newman, C. Moore: Finding community structure in very large
networks, Physical Review E 70.6 (Dec. 2004), issn: 1550-2376,
doi: 10.1103/physreve.70.066111,
url: http://dx.doi.org/10.1103/PhysRevE.70.066111.
[6] S. Fortunato: Community detection in graphs, Physics Reports 486.3 (2010), pp. 75–174,
issn: 0370-1573,
doi: https://doi.org/10.1016/j.physrep.2009.11.002.
[7] S. Fortunato, D. Hric: Community detection in networks: A user guide, Physics Reports
659 (2016), Community detection in networks: A user guide, pp. 1–44, issn: 0370-1573,
doi: https://doi.org/10.1016/j.physrep.2016.09.002.
137
[8] K. M. Ghori, R. A. Abbasi, M. Awais, M. Imran, A. Ullah, L. Szathmary: Per-
formance analysis of different types of machine learning classifiers for non-technical loss
detection, IEEE Access 8 (2019), pp. 16033–16048.
[9] K. M. Ghori, A. R. Ayaz, M. Awais, M. Imran, A. Ullah, L. Szathmary: Impact of
Feature Selection on Non-technical Loss Detection, in: 2020 6th Conference on Data Science
and Machine Learning Applications (CDMA), IEEE, 2020, pp. 19–24.
[10] K. M. Ghori, M. Imran, A. Nawaz, R. A. Abbasi, A. Ullah, L. Szathmary: Perfor-
mance analysis of machine learning classifiers for non-technical loss detection, Journal of
Ambient Intelligence and Humanized Computing (2020), pp. 1–16.
[11] J. I. Guerrero, I. Monedero, F. Biscarri, J. Biscarri, R. Millán, C. León: Non-
Technical Losses Reduction by Improving the Inspections Accuracy in a Power Utility, IEEE
Transactions on Power Systems 33.2 (2018), pp. 1209–1218.
[12] W. Han, Y. Xiao: NFD: a practical scheme to detect non-technical loss fraud in smart grid,
in: Communications (ICC), 2014 IEEE International Conference on, IEEE, 2014, pp. 605–
609.
[13] T. Hartmann, A. Moawad, F. Fouquet, Y. Reckinger, T. Mouelhi, J. Klein, Y.
Le Traon: Suspicious electric consumption detection based on multi-profiling using live
machine learning, in: Smart Grid Communications (SmartGridComm), 2015 IEEE Interna-
tional Conference on, IEEE, 2015, pp. 891–896.
[14] W. Hu, Y. Yang, J. Wang, X. Huang, Z. Cheng: Understanding Electricity-Theft Be-
havior via Multi-Source Data, arXiv preprint arXiv:2001.07311 (2020).
[15] L. A. P. Júnior, C. C. O. Ramos, D. Rodrigues, D. R. Pereira, A. N. de Souza,
K. A. P. da Costa, J. P. Papa: Unsupervised non-technical losses identification through
optimum-path forest, Electric Power Systems Research 140 (2016), pp. 413–423.
[16] H. Kwak, Y.-H. Eom, Y. Choi, H. Jeong, S. Moon: Consistent Community Identification
in Complex Networks, 2009, arXiv: 0910.1508 [physics.soc-ph].
[17] C. León, F. Biscarri, I. Monedero, J. I. Guerrero, J. Biscarri, R. Millán: Variabil-
ity and trend-based generalized rule induction model to NTL detection in power companies,
IEEE Transactions on Power Systems 26.4 (2011), pp. 1798–1807.
[18] J. A. Meira, P. Glauner, R. State, P. Valtchev, L. Dolberg, F. Bettinger, D.
Duarte: Distilling provider-independent data for general detection of non-technical losses,
in: Power and Energy Conference at Illinois (PECI), 2017 IEEE, IEEE, 2017, pp. 1–5.
[19] R. Mutupe, S. Osuri, M. Lencwe, S. D. Chowdhury: Electricity theft detection system
with RF communication between distribution and customer usage, in: PowerAfrica, 2017
IEEE PES, IEEE, 2017, pp. 566–572.
[20] M. E. J. Newman: Modularity and community structure in networks, Proceedings of the
National Academy of Sciences 103.23 (2006), pp. 8577–8582, issn: 0027-8424,
doi: 10.1073/pnas.0601602103, eprint: https://www.pnas.org/content/103/23/8577.
full.pdf,
url: https://www.pnas.org/content/103/23/8577.
[21] M. E. J. Newman, M. Girvan: Finding and evaluating community structure in networks,
Phys. Rev. E 69 (2 Feb. 2004), p. 026113,
doi: 10.1103/PhysRevE.69.026113,
url: https://link.aps.org/doi/10.1103/PhysRevE.69.026113.
[22] G. Palla, I. Derényi, I. Farkas, T. Vicsek: Uncovering the overlapping community
structure of complex networks in nature and society, Nature 435 (2005), pp. 814–818, issn:
1476-4687,
doi: 10.1038/nature03607,
url: https://doi.org/10.1038/nature03607.
138
[23] S. Pazi, C. M. Clohessy, G. D. Sharp: A framework to select a classification algorithm
in electricity fraud detection, en, South African Journal of Science 116 (Oct. 2020), pp. 1–7,
issn: 0038-2353,
url: http : / / www . scielo . org . za / scielo . php ? script = sci _ arttext & pid = S0038 -
23532020000600010&nrm=iso.
[24] B. Peng, C. Wan, S. Dong, J. Lin, Y. Song, Y. Zhang, J. Xiong: A two-stage pat-
tern recognition method for electric customer classification in smart grid, in: Smart Grid
Communications (SmartGridComm), 2016 IEEE International Conference on, IEEE, 2016,
pp. 758–763.
[25] Q. A. Al-Radaideh, M. M. Al-Zoubi: A data mining based model for detection of fraud-
ulent behaviour in water consumption, in: 2018 9th International Conference on Information
and Communication Systems (ICICS), IEEE, 2018, pp. 48–54.
[26] C. C. Sánchez-Zuleta, J. P. Fernández-Gutiérrez, C. C. Piedrahita-Escobar: Iden-
tification of the characteristics incident to the detection of non-technical losses for two
Colombian energy companies, Revista Facultad de Ingeniería Universidad de Antioquia 84
(2017), pp. 60–71.
[27] D. D. Sharma, S. Singh: Aberration detection in electricity consumption using clustering
technique, International Journal of Energy Sector Management 9.4 (2015), pp. 451–470.
[28] D. D. Sharma, S. Singh, L. Jeremy, E. Foruzan: Identification and characterization of
irregular consumptions of load data, Journal of Modern Power Systems and Clean Energy
5.3 (2017), pp. 465–477.
[29] J. V. Spirić, S. S. Stanković, M. B. Dočić: Identification of suspicious electricity cus-
tomers, International Journal of Electrical Power & Energy Systems 95 (2018), pp. 635–
643.
[30] M. A. Tahir, J. Kittler, K. Mikolajczyk, F. Yan: A multiple expert approach to the
class imbalance problem using inverse random under sampling, in: International workshop
on multiple classifier systems, Springer, 2009, pp. 82–91.
[31] E. Terciyanli, E. Eryigit, T. Emre, S. Caliskan: Score based non-technical loss detec-
tion algorithm for electricity distribution networks, in: Smart Grid and Cities Congress and
Fair (ICSG), 2017 5th International Istanbul, IEEE, 2017, pp. 180–184.
[32] J. L. Viegas, P. R. Esteves, S. M. Vieira: Clustering-based novelty detection for identi-
fication of non-technical losses, International Journal of Electrical Power & Energy Systems
101 (2018), pp. 301–310.
[33] J. Yeckle, B. Tang: Detection of Electricity Theft in Customer Consumption Using Out-
lier Detection Algorithms, in: Data Intelligence and Security (ICDIS), 2018 1st International
Conference on, IEEE, 2018, pp. 135–140.
[34] K. Zheng, Y. Wang, Q. Chen, Y. Li: Electricity theft detecting based on density-clustering
method, in: 2017 IEEE Innovative Smart Grid Technologies - Asia (ISGT-Asia), 2017, pp. 1–
6,
doi: 10.1109/ISGT-Asia.2017.8378347.
[35] K. Zheng, Y. Wang, Q. Chen, Y. Li: Electricity theft detecting based on density-clustering
method, in: Innovative Smart Grid Technologies-Asia (ISGT-Asia), 2017 IEEE, IEEE, 2017,
pp. 1–6.
[36] Z. Zheng, Y. Yang, X. Niu, H.-N. Dai, Y. Zhou: Wide and deep convolutional neural
networks for electricity-theft detection to secure smart grids, IEEE Transactions on Industrial
Informatics 14.4 (2018), pp. 1606–1615.
139