Artificial Intelligence techniques in Chronic

condition Lung Disease

ABSTRACT:
Background: Artificial Intelligence (AI) has proven to be an invaluable
asset in the healthcare domain, where massive amounts of data are produced.
Chronic Obstructive Pulmonary Disease (COPD) is a heterogeneous chronic
condition with multiscale manifestations and complex interactions that
represents an ideal target for AI. Objective: The aim of this review article is to
appraise the adoption of AI in COPD research, and more specifically its
applications to date along with reported results, potential challenges and future
prospects. Methods: We performed a review of the literature from PubMed and
DBLP and assembled studies published up to 2020, yielding 156 articles
relevant to the scope of this review. Results: The resulting articles were
assessed and organized into four basic contextual categories, namely: i) COPD
diagnosis, ii) COPD prognosis, iii) Patient classification, iv) COPD
management, and subsequently presented in an orderly manner based on a set of
qualitative and quantitative criteria. Conclusions: We observed considerable
acceleration of research activity utilizing AI techniques in COPD research,
especially in the last couple of years, nevertheless, the massive production of
large and complex data in COPD calls for broader adoption of AI and more
advanced techniques.
Keyword: Balanced probability distribution (BPD) algorithm, chronic
obstructive pulmonary disease (COPD), feature extraction, few-shot learning,
transfer learning
INTRODUCTION:
Chronic obstructive pulmonary disease (COPD) is a preventable and treatable
common disease characterized by persistent respiratory symptoms and restricted
airflow. According to the global initiative for COPD (hereinafter GOLD), this
sickness is often associated with airway or alveolar abnormalities caused by
significant exposure to toxic particles or gases. COPD is a chronic respiratory
disease that seriously threatens people’s health and has a high incidence and
mortality globally. The World Health Organization has reported that COPD has
become the third leading cause of deaths globally and will become one of the
leading respiratory diseases in China by 2020 [1]. In China, the prevalence of
COPD in people over 40 years of age is 9.9%. The morbidity and mortality
associated with COPD are often underestimated because of the differences in
the diagnostic criteria and an insufficient understanding of the disease; the rate
of missed diagnosis is as high as 70% [2]. Therefore, a predictive data mining
model with good clinical reliability is an important requirement for diagnosis,
treatment, and self-management of COPD [3]. In recent years, different
machine-learning models have been used for predicting COPD [1], which can
be characterized as either non-deep (i.e., traditional) or deep [4]. A traditional
model typically comprises two major steps: feature engineering [5] and model
building [6]. Feature engineering extracts the ‘‘good’’ features that are effective
for constructing the model. Unlike traditional methods, the deep learning model
[7] has an end-to-end learning mechanism in which the feature engineering part
is implicitly integrated into the learning pipeline. Deep learning has attracted the
attention formance. In deep learning, it is possible to learn advanced features
using big data; therefore, it is a representationlearning algorithm based on large-
scale data used in machine learning. However, in some bioinformatics fields,
such as COPD diagnosis, it is very difficult to construct large-scale well-labeled
datasets because of the high cost of data collection and labeling; this limits the
development of deep learning in these fields. If only a small amount of data is
used for training, it will lead to the problem of model overfitting and low
reliability [8]. Moreover, imbalances in medical data make it difficult to obtain
efficient disease prediction models. Transfer learning [9] relaxes the assumption
that training data must be independent and identically distributed (i.i.d) with the
test data. In addition, transfer learning applies the knowledge learned from
existing tasks to new models or fields to improve the learning performance and
obtain efficient prediction models greatly. In fact, transfer learning has become
a new learning framework to solve many knowledge transfer problems [10]. In
this work, it has a great positive impact on performance improvement despite
limited training data. There have been numerous studies on COPD [11] since
the 1980s. However, the existing treatment methods for COPD do not consider
the possible complications and lack a systematic approach to exploring the
interactions between COPD and the complications. Clinically, the pathogenesis,
clinical manifestations, diagnosis, treatment, and management of COPD
complications and simple COPD+ complications will have varying degrees of
differences. As stated by GOLD in 2017, the natural course of COPD develops
from complex systematic outcomes and complications, which are the main
characteristics of clinical COPD [1]. The systemic outcome of COPD refers to
the direct pulmonary phenotype and the non-pulmonary manifestations caused
by COPD.
LITERATURE SURVEY:
1.TOPIC: Global strategy for the diagnosis, management, and prevention of
chronic obstructive lung disease 2017 Report. GOLD executive summary
AUTHOR: C. F. Vogelmeier et al.,
This Executive Summary of the Global Strategy for the Diagnosis,
Management, and Prevention of COPD (GOLD) 2017 Report focuses primarily
on the revised and novel parts of the document. The most significant changes
include: 1) the assessment of chronic obstructive pulmonary disease has been
refined to separate the spirometric assessment from symptom evaluation. ABCD
groups are now proposed to be derived exclusively from patient symptoms and
their history of exacerbations; 2) for each of the groups A to D, escalation
strategies for pharmacological treatments are proposed; 3) the concept of de-
escalation of therapy is introduced in the treatment assessment scheme; 4)
nonpharmacologic therapies are comprehensively presented and; 5) the
importance of comorbid conditions in managing COPD is reviewed.
2.TOPIC: Prevalence and underdiagnosis of airway obstruction among middle-
aged adults in Northern France: The ELISABET study 2011–2013
AUTHOR: A. Quach, J. Giovannelli, N. Chérot-Kornobis, A. Ciuchete, G.
Clément, R. Matran, P. Amouyel, J.-L. Edmé, and L. Dauchet
Airway obstruction (AO), mainly due to chronic obstructive pulmonary
disease (COPD) in adults, is a major cause of mortality and poor quality of life.
However, few data are available for France. This study was designed to
calculate the prevalence AO among middle-aged adults in northern France,
explore the associated risk factors and evaluate the underdiagnosis. The Enquête
Littoral Souffle Air Biologie Environnement (ELISABET) was a cross-sectional
study of a representative sample of 3276 adults aged from 40 to 64 in two urban
areas in northern France (Lille and Dunkirk). Participants filled out a
questionnaire and performed spirometry testing, without a reversibility test. The
age-standardized estimated prevalence [95% confidence interval] of AO was
16.0% [13.9; 17.9] in Lille and 13.7% [11.7; 15.7] in Dunkirk with the Global
initiative for chronic Obstructive Lung Disease (GOLD) definition and 10.8%
[9.2; 12.5] and 9.5% [7.9; 11.2] respectively with the lower limit of normal
calculated with the Global Lung Initiative (GLI) 2012 equations. AO was
associated with age, male gender, tobacco consumption and low body mass
index. The under diagnosis rate was greater than 70%. Previously undiagnosed
participants with AO displayed more respiratory symptoms compared with
participants without AO and less than participants with previously diagnosed
AO.
3. TOPIC: Use of predictive algorithms in-home monitoring of chronic
obstructive pulmonary disease and asthma: A systematic review
AUTHOR: D. Sanchez-Morillo, M. A. Fernandez-Granero, and A. Leon-
Jimenez
 Major reported factors associated with the limited effectiveness of home
tele monitoring interventions in chronic respiratory conditions include the lack
of useful early predictors, poor patient compliance and the poor performance of
conventional algorithms for detecting deteriorations. This article provides a
systematic review of existing algorithms and the factors associated with their
performance in detecting exacerbations and supporting clinical decisions in
patients with chronic obstructive pulmonary disease (COPD) or asthma. An
electronic literature search in Medline, Scopus, Web of Science and Cochrane
library was conducted to identify relevant articles published between 2005 and
July 2015. A total of 20 studies (16 COPD, 4 asthma) that included research
about the use of algorithms in tele monitoring interventions in asthma and
COPD were selected. Differences on the applied definition of exacerbation, tele
monitoring duration, acquired physiological signals and symptoms, type of
technology deployed and algorithms used were found. Predictive models with
good clinically reliability have yet to be defined, and are an important goal for
the future development of tele health in chronic respiratory conditions. New
predictive models incorporating both symptoms and physiological signals are
being tested in tele monitoring interventions with positive outcomes. However,
the underpinning algorithms behind these models need be validated in larger
samples of patients, for longer periods of time and with well-established
protocols. In addition, further research is needed to identify novel predictors
that enable the early detection of deteriorations, especially in COPD. Only then
will tele monitoring achieve the aim of preventing hospital admissions,
contributing to the reduction of health resource utilization and improving the
quality of life of patients
4. TOPIC: Predictive modeling of the hospital readmission risk from Patients’
claims data using machine learning: A case study on COPD,
AUTHOR: X. Min, B. Yu, and F. Wang
 Heart failure hospitalization is a severe burden on healthcare. How to
predict and therefore prevent readmission has been a significant challenge in
outcomes research. To address this, we propose a deep learning approach to
predict readmission from clinical notes. Unlike conventional methods that use
structured data for prediction, we leverage the unstructured clinical notes to
train deep learning models based on convolutional neural networks (CNN). We
then use the trained models to classify and predict potentially high-risk
admissions/patients. For evaluation, we trained CNNs using the discharge
summary notes in the MIMIC III database. We also trained regular machine
learning models based on random forest using the same datasets. The result
shows that deep learning models outperform the regular models in prediction
tasks. CNN method achieves a F1 score of 0.756 in general readmission
prediction and 0.733 in 30-day readmission prediction, while random forest
only achieves a F1 score of 0.674 and 0.656 respectively. We also propose a
chi-square test based method to interpret key features associated with deep
learning predicted readmissions. It reveals clinical insights about readmission
embedded in the clinical notes. Collectively, our method can make the human
evaluation process more efficient and potentially facilitate the reduction of
readmission rates.
5. TOPIC:Shared-nearest-neighbor-based clustering by fast search and find of
density peaks,
AUTHOR: R. Liu, H. Wang, and X. Yu
Clustering by fast search and find of density peaks (DPC) is a
new clustering method that was reported in Science in June 2014.
This clustering algorithm is based on the assumption that cluster centers have
high local densities and are generally far from each other. With a decision
graph, cluster centers can be easily located. However, this approach suffers
from certain disadvantages. First, the definition of the local density and distance
measurement is too simple; therefore, the DPC algorithm might perform poorly
on complex datasets that are of multiple scales, cross-winding, of various
densities, or of high dimensionality. Second, the one-step allocation strategy is
not robust and has poor fault tolerance. Thus, if a point is assigned incorrectly,
then the subsequent allocation will further amplify the error, resulting in more
errors, which will have a severe negative impact on the clustering results. Third,
the cutoff distance dc is generally difficult to determine since the range of each
attribute is unknown in most cases. Even when being normalized or using the
relative percentage method, a small change in dc will still cause a conspicuous
fluctuation in the result, and this is especially true for real-world datasets.
Considering these drawbacks, we propose a shared-nearest-neighbor-based
clustering by fast search and find of density peaks (SNN-DPC) algorithm. We
present three new definitions: SNN similarity, local density ρ and distance from
the nearest larger density point δ. These definitions take the information of the
nearest neighbors and the shared neighbors into account, and they can self-adapt
to the local surroundings. Then, we introduce our two-step allocation method:
inevitably subordinate and possibly subordinate. The former quickly and
accurately recognizes and allocates the points that certainly belong to one
cluster by counting the number of shared neighbors between two points. The
latter assigns the remaining points by finding the clusters to which more
neighbors belong. The algorithm is benchmarked on publicly available synthetic
datasets, UCI real-world datasets and the Olivetti Faces dataset, which are often
used to test the performance of clustering algorithms. We compared the results
with those of DPC, fuzzy weighted K-nearest neighbors density peak clustering
(FKNN-DPC), affinity propagation (AP), ordering points to identify
the clustering structure (OPTICS), density-based spatial clustering of
applications with noise (DBSCAN), and K-means. The metrics used are
adjusted mutual information (AMI), adjusted Rand index (ARI), and Fowlkes–
Mallows index (FMI). The experimental results prove that our method can
recognize clusters regardless of their size, shape, and dimensions; is robust to
noise; and is remarkably superior to DPC, FKNN-DPC, AP, OPTICS,
DBSCAN, and K-means.
EXISTING SYSTEM:
The training data and test data with i.i.d are the necessary requirements for a
training model that has good prediction accuracy in the test data. However,
uniformly distributed data are rare in real life. For example, if a model trained
on book review texts is directly applied for predicting movie review text data, it
is likely that the result would be unsatisfactory because of the different data
distributions. Therefore, training models with sound testing effects in multiple
fields are particularly important for practical situations. Transfer learning can
determine the potential relationship between the source and target domains and
further build a suitable target domain model based on the learned source domain
knowledge. Therefore, narrowing the data distribution distance between the
fields is the key to cross-domain learning and building models. However, the
data from different domains have different distributions, which makes it
difficult to fit all the distributions simultaneously. Moreover, simply matching
the distributions cannot guarantee the prediction effect on the target domain.
Therefore, in this paper, we propose the BPD algorithm based on the instance
and feature transfer methods. The BPD framework

PROPOSED SYSTEM:

CONCLUSION:
  We observed considerable acceleration of research activity utilizing AI
techniques in COPD research, especially in the last couple of years,
nevertheless, the massive production of large and complex data in COPD calls
for broader adoption of AI and more advanced techniques.

REFERENCE:
[1] C. F. Vogelmeier et al., ‘‘Global strategy for the diagnosis, management,
and prevention of chronic obstructive lung disease 2017 Report. GOLD
executive summary,’’ Amer. J. Respiratory Crit. Care Med., vol. 195, no. 5, pp.
557–582, Mar. 2017.
[2] A. Quach, J. Giovannelli, N. Chérot-Kornobis, A. Ciuchete, G. Clément, R.
Matran, P. Amouyel, J.-L. Edmé, and L. Dauchet, ‘‘Prevalence and
underdiagnosis of airway obstruction among middle-aged adults in Northern
France: The ELISABET study 2011–2013,’’ Respiratory Med., vol. 109, no. 12,
pp. 1553–1561, Dec. 2015.
[3] D. Sanchez-Morillo, M. A. Fernandez-Granero, and A. Leon-Jimenez, ‘‘Use
of predictive algorithms in-home monitoring of chronic obstructive pulmonary
disease and asthma: A systematic review,’’ Chronic Respiratory Disease, vol.
13, no. 3, pp. 264–283, Aug. 2016.
[4] X. Min, B. Yu, and F. Wang, ‘‘Predictive modeling of the hospital
readmission risk from Patients’ claims data using machine learning: A case
study on COPD,’’ Sci. Rep., vol. 9, no. 1, pp. 1–10, Dec. 2019.
[5] R. Liu, H. Wang, and X. Yu, ‘‘Shared-nearest-neighbor-based clustering by
fast search and find of density peaks,’’ Inf. Sci., vol. 450, pp. 200–226, Jun.
2018. [6] B. Hu, H. Wang, X. Yu, W. Yuan, and T. He, ‘‘Sparse network
embedding for community detection and sign prediction in signed social
networks,’’ J. Ambient Intell. Hum. Comput., vol. 10, no. 1, pp. 175–186, Jan.
2019.
[7] Y. LeCun, Y. Bengio, and G. Hinton, ‘‘Deep learning,’’ Nature, vol. 521,
no. 7553, pp. 436–444, 2015.
[8] F. Shen, J. Chao, and J. Zhao, ‘‘Forecasting exchange rate using deep belief
networks and conjugate gradient method,’’ Neurocomputing, vol. 167, pp. 243–
253, Nov. 2015.
[9] S. J. Pan and Q. Yang, ‘‘A survey on transfer learning,’’ IEEE Trans.
Knowl. Data Eng., vol. 22, no. 10, pp. 1345–1359, Oct. 2010.
[10] G. Wang, J. Qiao, J. Bi, W. Li, and M. Zhou, ‘‘TL-GDBN: Growing deep
belief network with transfer learning,’’ IEEE Trans. Autom. Sci. Eng., vol. 16,
no. 2, pp. 874–885, Apr. 2019
[11] G. Lu, D. Li, and L. Zhang, ‘‘Clinical investigation of depression in elderly
patients with chronic obstructive pulmonary disease,’’ China Med. Pharmacy,
vol. 3, no. 1, pp. 12–14, 2013.
[12] J. Bi, T. Feng, and H. Yuan, ‘‘Real-time and short-term anomaly detection
for GWAC light curves,’’ Comput. Ind., vol. 97, pp. 76–84, May 2018.
[13] T. S. Brisimi, T. Xu, T. Wang, W. Dai, W. G. Adams, and I. C.
Paschalidis, ‘‘Predicting chronic disease hospitalizations from electronic health
records: An interpretable classification approach,’’ Proc. IEEE, vol. 106, no. 4,
pp. 690–707, Apr. 2018.
[14] L. Han, M. Maciejewski, C. Brockel, W. Gordon, S. B. Snapper, J. R.
Korzenik, L. Afzelius, and R. B. Altman, ‘‘A probabilistic pathway score
(PROPS) for classification with applications to inflammatory bowel disease,’’
Bioinformatics, vol. 34, no. 6, pp. 985–993, Mar. 2018.
[15] A. Yosipof, R. C. Guedes, and A. T. García-Sosa, ‘‘Data mining and
machine learning models for predicting drug likeness and their disease or organ
category,’’ Frontiers Chem., vol. 6, p. 162, May 2018.
[16] Z. C. Lipton, D. C. Kale, C. Elkan, and R. Wetzel, ‘‘Learning to diagnose
with LSTM recurrent neural networks,’’ 2015, arXiv:1511.03677. [Online].
Available: http://arxiv.org/abs/1511.03677
[17] M. Anthimopoulos, S. Christodoulidis, L. Ebner, A. Christe, and S.
Mougiakakou, ‘‘Lung pattern classification for interstitial lung diseases using a
deep convolutional neural network,’’ IEEE Trans. Med. Imag., vol. 35, no. 5,
pp. 1207–1216, May 2016.
[18] X. Li, H. Wang, Y. Xiong, Y. Zeng, T. He, J. Wu, and J. Chen,
‘‘Application of convolutional neural network in pediatric disease prediction,’’
China Digit. Med., vol. 13, no. 10, pp. 11–13, 2018.
[19] E. Prescott, P. Lange, and J. Vestbo, ‘‘Socioeconomic status, lung function
and admission to hospital for COPD: Results from the Copenhagen city heart
study,’’ Eur. Respiratory J., vol. 13, no. 5, pp. 1109–1114, May 1999.
[20] M. Kanervisto, T. Vasankari, T. Laitinen, M. Heliövaara, P. Jousilahti, and
S. Saarelainen, ‘‘Low socioeconomic status is associated with chronic
obstructive airway diseases,’’ Respiratory Med., vol. 105, no. 8, pp. 1140–1146,
Aug. 2011.
[21] P. Danielsson, I. S. Ólafsdóttir, B. Benediktsdóttir, T. Gíslason, and C.
Janson, ‘‘The prevalence of chronic obstructive pulmonary disease in Uppsala,
Sweden—The burden of obstructive lung disease (BOLD) study: Cross-
sectional population-based study,’’ Clin. Respiratory J., vol. 6, no. 2, pp. 120–
127, Apr. 2012.
[22] Y. Borné, W. Ashraf, S. Zaigham, and S. Frantz, ‘‘Socioeconomic
circumstances and incidence of chronic obstructive pulmonary disease (COPD)
in an urban population in sweden,’’ COPD: J. Chronic Obstructive Pulmonary
Disease, vol. 16, no. 1, pp. 51–57, Jan. 2019.
[23] H. Zhao et al., ‘‘Relationship between stress hormone levels and prognosis
in elderly patients with COPD,’’ J. Guangxi Univ. Traditional Chin. Med., vol.
22, no. 2, pp. 26–28, 2019.
[24] J. M. Marin, S. J. Carrizo, C. Casanova, P. Martinez-Camblor, J. B.
Soriano, A. G. N. Agusti, and B. R. Celli, ‘‘Prediction of risk of COPD
exacerbations by the BODE index,’’ Respiratory Med., vol. 103, no. 3, pp. 373–
378, Mar. 2009. [25] M. H. Jensen, S. L. Cichosz, B. Dinesen, and O. K.
Hejlesen, ‘‘Moving prediction of exacerbation in chronic obstructive pulmonary
disease for patients in telecare,’’ J. Telemed. Telecare, vol. 18, no. 2, pp. 99–
103, Mar. 2012.
[26] M. van der Heijden, P. J. F. Lucas, B. Lijnse, Y. F. Heijdra, and T. R. J.
Schermer, ‘‘An autonomous mobile system for the management of COPD,’’ J.
Biomed. Informat., vol. 46, no. 3, pp. 458–469, Jun. 2013.
[27] C. Burton, H. Pinnock, and B. McKinstry, ‘‘Changes in telemonitored
physiological variables and symptoms prior to exacerbations of chronic
obstructive pulmonary disease,’’ J. Telemed. Telecare, vol. 21, no. 1, pp. 29–36,
Jan. 2015.
[28] W. Dai, Q. Yang, G.-R. Xue, and Y. Yu, ‘‘Boosting for transfer learning,’’
in Proc. 24th Int. Conf. Mach. Learn. (ICML), 2007, pp. 193–200.
[29] S. J. Pan, I. W. Tsang, J. T. Kwok, and Q. Yang, ‘‘Domain adaptation via
transfer component analysis,’’ IEEE Trans. Neural Netw., vol. 22, no. 2, pp.
199–210, Feb. 2011.
[30] K. M. Borgwardt, A. Gretton, M. J. Rasch, H.-P. Kriegel, B. Schölkopf,
and A. J. Smola, ‘‘Integrating structured biological data by kernel maximum
mean discrepancy,’’ Bioinformatics, vol. 22, no. 14, pp. e49–e57, Jul. 2006.
[31] M. M. Ohayon, ‘‘Chronic obstructive pulmonary disease and its
association with sleep and mental disorders in the general population,’’ J.
Psychiatric Res., vol. 54, pp. 79–84, Jul. 2014. [32] N. Zhong and C. Cai,
‘‘Relationship between chronic obstructive pulmonary disease and anxiety
depression,’’ Cont. Med. Educ., vol. 21, no. 16, pp. 17–19, 2006.
[33] Y. Zhu and W. Zhang, ‘‘Discussion on the treatment of chronic obstructive
pulmonary disease from lung, spleen and kidney,’’ Shanxi J. Traditional Chin.
Med., vol. 31, no. 7, pp. 1–2, 2015.
[34] J. Bi, H. Yuan, M. Zhou, and Q. Liu, ‘‘Time-dependent cloud workload
forecasting via multi-task learning,’’ IEEE Robot. Autom. Lett., vol. 4, no. 3,
pp. 2401–2406, Jul. 2019. [35] X. Guo, M. Zhou, S. Liu, and L. Qi,
‘‘Lexicographic multiobjective scatter search for the optimization of sequence-
dependent selective disassembly subject to multiresource constraints,’’ IEEE
Trans. Cybern., to be published.
[36] X. Luo, M. Zhou, Z. Wang, Y. Xia, and Q. Zhu, ‘‘An effective scheme for
QoS estimation via alternating direction method-based matrix factorization,’’
IEEE Trans. Services Comput., vol. 12, no. 4, pp. 503–518, Jul./Aug. 2019.
[37] R. Veevers and S. Hayward, ‘‘Morphing and docking visualisation of
biomolecular structures using multi-dimensional scaling,’’ J. Mol. Graph.
Model., vol. 82, pp. 108–116, Jun. 2018.
[38] L. Yu and H. Liu, ‘‘Efficient feature selection via analysis of relevance
andredundancy,’’ J. Mach. Learn. Res., vol. 5, no. Oct, pp. 1205–1224, 2004.
[39] K. Passi, A. Nour, and C. K. Jain, ‘‘Markov blanket: Efficient strategy for
feature subset selection method for high dimensional microarray cancer
datasets,’’ in Proc. IEEE Int. Conf. Bioinf. Biomed. (BIBM), Nov. 2017, pp.
1864–1871.
[40] J. Bi, H. Yuan, and M. Zhou, ‘‘Temporal prediction of multiapplication
consolidated workloads in distributed clouds,’’ IEEE Trans. Autom. Sci. Eng.,
vol. 16, no. 4, pp. 1763–1773, Oct. 2019.
[41] H. Wang, A. Kembhavi, A. Farhadi, A. L. Yuille, and M. Rastegari,
‘‘ELASTIC: Improving CNNs with dynamic scaling policies,’’ in Proc.
IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), Jun. 2019, pp. 2258–
2267.
[42] Y. Fang, H. Wang, L. Wang, R. Di, and Y. Song, ‘‘Feature-maximum
dependency-based fusion diagnosis method for COPD,’’ Multimedia Tools
Appl., vol. 78, no. 1, pp. 1–18, 2019.
[43] G. Wang, Q.-S. Jia, J. Qiao, J. Bi, and C. Liu, ‘‘A sparse deep belief
network with efficient fuzzy learning framework,’’ Neural Netw., vol. 121, pp.
430–440, Jan. 2020.
[44] J. Bi, H. Yuan, L. Zhang, and J. Zhang, ‘‘SGW-SCN: An integrated
machine learning approach for workload forecasting in geo-distributed cloud
data centers,’’ Inf. Sci., vol. 481, pp. 57–68, May 2019.