Advancing Manufacturing with Machine Learning:

Unlocking the Potential of Reinforcement learning in

Industry 4.0
Zineb Ihrouchen, Omar Souissi, Hicham Bensaid

To cite this version:

Zineb Ihrouchen, Omar Souissi, Hicham Bensaid. Advancing Manufacturing with Machine Learning:
Unlocking the Potential of Reinforcement learning in Industry 4.0. Industrial Conference of Science
and Industry of the Future (ICSIF23), Oct 2023, Rabat, Morocco. �hal-04284006�

HAL Id: hal-04284006

https://hal.science/hal-04284006
Submitted on 14 Nov 2023

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est

archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents
entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non,
lished or not. The documents may come from émanant des établissements d’enseignement et de
teaching and research institutions in France or recherche français ou étrangers, des laboratoires
abroad, or from public or private research centers. publics ou privés.
 Advancing Manufacturing with Machine Learning:
Unlocking the Potential of Reinforcement learning in Industry 4.0

Zineb IHROUCHEN1 , Omar SOUISSI2 , and Hicham BENSAID3

1
Data Analytic, operations and decision, INPT Rabat, Morocco, zineb.ihrouchen@gmail.com
2
Data Analytic, operations and decisions, INPT Rabat, Morocco, souissi@inpt.ac.ma
3
Smart, Embedded, Enterprise and Distributed Systems , INPT Rabat, Morocco, bensaid@inpt.ac.ma

Abstract. The ongoing Industry 4.0 revolution is transforming manufacturing environment to be

smarter, empowering machines to comprehend processes, interact with their environment, and adapt
intelligently, thus enhancing the productivity of manu- facturing technologies. Within the realm of AI
(Artificial In- telligence), Machine Learning (ML) has emerged as a pivotal force propelling these in-
novations in various industrial sectors, and leading digital transformation of operations and processes
to Industry of the future. Particularly, ML, with a focus on Reinforcement Learning (RL), has demon-
strated its prowess in successfully addressing diverse tasks and challenges across different complexities
and performance levels in the industry. This paper provides a comprehensive overview of ML applica-
tions for smart manufacturing, highlighting several powerful ML algorithms capable of offering solutions
to pertinent issues while averting potential challenges in the future of the industry.

Keywords: Industry 4.0 · Machine Learning · Reinforcement learning · artificial intelligence

1 Introduction

Over last two centuries, every industrial revolution has been propelled by a change in production methods
through an emerging technology of that era. In this context, the steam engine was the flame that started
the first revolution in industry, the second started by the application of electricity, and the third by the
widespread use of information technology and electronics. In recent times, the popularity of the Internet of
Things (IoT) [1] and Cyber- Physical Systems (CPS) has captured the attention of entreprise and academia.
The use of these technologies promises to enable better connectivity between information, products, and peo-
ple, thereby improving current production models. This phenomenon is known as Industry 4.0 [2]. It leads to
several innovations, from digital technologies, advanced materials, innovative products to new manufacturing
processes. Industry 4.0 is a revolutionary concept that is having a major impact on several areas of industrial
manufacturing . It helps to enhance and improve development and growth through the use and adaptation
of new technology and engineering tools, for creating intelligent products through intelligent processes. This
evolution in industry is realised including various technologies such as : ML, Internet of Things, Big Data, and
Cloud Computing (CC) which are driving together a revolutionary transformation in the fields of industry .
Smart manufacturing, a key aspect of Industry 4.0, epitomize the forthcoming paradigm of manufacturing,
wherein the uninterrupted transmission and meticulous analysis of shop floor data play a pivotal role in gen-
erating profound manufacturing insights. In this context of industry of the future a wide range of applications
is envisioned, providing a solution that can help manufacturing companies to further optimise their processes
and significantly increase their internal efficiency, thanks to the system’s automated procedures, easy setup
including simple, demand-driven installation and, finally, its high degree of scalability. Depending on theses
application, a various of ML algorithms, have already been explored to achieve characteristics such as fault
detection, prediction and autonomy.
AI and ML, gave a tremendous boost to Industry 4.0. Thanks to intelligent software, the large volumes of
information and data being collected from the industrial environment through IoT devices. AI solutions can
learn to read ”between the lines”, enabling the discovery of many complex interrelationships that human
beings would not or no longer be able to grasp, this enhanced capability is made possible through the use of
advanced ML analysis techniques. The increasing amount of real-time data generated in industrial environ-
ment has led factories to the adopt ML’s techniques, algorithmes to respond specific industrial needs.
According to Arthur L. Samuel [35], ML refers to the ability of a system to behave in a smart way like
human, in terms of performance and solving complex and hard problems. There are different types of ML
models, which are categorised[36] as follows: supervised learning (include algorithms of classification and
regression such as Linear Discriminant Analysis (LDA) and Linear Regression(LR)), unsupervised learn-
ing (include algorithms of clustering such as K-Means), reinforcement learning (RL) (is based on agent,
include algorithm such as Deep Q-learning, AlphaGo [37] is an example).
Recently ML applications have known a large expansion Not surprisingly, many works point to ML being the
key to the development of the present industry towards industry 4.0 evolution .[38][39] The rest of this paper
focus in section 2 on the methodology used to identify pertinent research and mention some related work to
the use of machine learning in industry 4.0 . The section 3 contains a definition of Reinforcement learning
approach and their application in industry 4.0 including some related work.

2 RELATED WORK
In this part, we discuss our used methodology and then we introduce some state of art in machine learning
applications for industry 4.0 between 2018 and 2023 according to several criteria. First, we start with ML
algorithms used in manufac- turing, then we mention their application in industry.

2.1 Used Methodology

In this part, we describe the methodology used to identify pertinent studies within the field of machine learning
in the context of industry 4.0. We conducted thorough searches across Google Scholar database , focusing
on English-language studies published between 2018 and 2023. Our search terms comprised a combination of
keywords, including ”ML/RL industrie 4.0”, ”industry of the future”, ”ML/RL smart industry” ,”ML/RL
scheduling industry,” ”ML/RL computer vision industry” ,”ML/RL predictive maintenance industry” and
”ML/RL could/edge computing industry”. Our approach entailed scrutinizing titles and abstracts for the
designated terms and additionally cross-referencing the complete bibliographies of located articles. We aimed
for a comprehensive research process, culminating in a curated selection of articles that objectively encapsulate
the existing literature on Machine learning and Reinforcement learning in industry of the future.
Most of articles included are recent publications (Figure 1.a). This reflects the trend of a sharp a sharp
growth of articles published over recents years; Google Trends illustrated in Figure 1.b prove an increase
interest score for Machine Learning and Industry 4.0 over recent 5 years.

(a) ML publications in industry 4.0 in the past five years (b) Machine Learning and industry 4.0 score Interest

Fig. 1: Machine learning application and Interest

2.2 Previous Survey Works

This part of paper focuses on giving a broad review of some ML algorithms employed in manufacturing
processes. In order to carry out succesfully our work , we are based on a systematic review (SR) approch
described in [4] to identify relevant research and work on ML techniques[5] and application in industry 4.0. In
recent times, the field of study related to ML applied in Industry 4.0 has attracted many contributions. The
paper in [6][7] explained Successful machine learning algorithms in manufacturing including artificial neural
networks(ANN) , support vector machines (SVM), Naive Bayesian (NB) , LR, and other ensemble methods.
Convolutional Neural Network (CNN) and Recurrent Neural Network (RNN) as example of deep learning al-
gorithms have known recently good efficiency in terms of performance in areas involving images recognation,
object detection and text processing/analysis [8]. The authors of Reference [9] propose for solving a produc-
tion chain problem in manufacturing systems, the use of a machine learning method named Classification
And Regression Tree (CART). The proposed approach used a decision tree based on supervised learning, to
optimize delivery costs. Furthermore, an overview of a defect detection systems based on machine learning
algorithm including a series of new digital manufacturing technology; was given in Reference [10]. A novel
approach for fault bearing diagnosis was introduced in Reference [11], it based on a deep belief network
(DBN) combining with Joint Distribution Adaptation (JDA) . This latest, plays a crucial role in the Joint
Distribution Adaptation with DBN (JACADN) method, allowing feature transfer between samples from the
source and target domains.
According to predictive maintenance, the work described in [12] proposes a model combining CNN and
LSTM the first refer to convolutional neural networks, the second for long short-term memory it is a forecat-
ing approch for predictive maintenance based on historical data Where the autors experiment the prediction
of machine failures. Moreover, the Reference [13] provide a predictive and preventive maintenance model
applicable to the industrial field using RNN. A stochastic dynamic programming solving approach is pre-
sented by authors in Reference [31] for solving the problem of machine’s breakdowns, by scheduling jobs and
implementing preventive maintenance, aiming to minimize the cost associated with earliness and tardiness.
The study in the paper in Refrence [33] presents experimental findings that highlight the adaptive large
neighborhood search algorithm’s (ALNS) competitive performance to solve Hard scheduling problem with
different specific time. It demonstrates a capacity to match or surpass the state of the art in 63% of cases.
The aim of the study in reference[14] is to review several works on ML applications for the optimization pro-
duction processes and for the improvement of products quality. in that regard, a method used for inspecting
a quality of a product in manufacturing environment is proposed in Reference[15]. The model of inspection
of quality product realised in this approach, is based on selecting data, then checking and controlling the
quality, finally training the model to be adaptive to different cases, using supervised ML algorithms including
Gradient Boosted Tree (GBT), Logistic Regression (LR) also , Support Vector Machines (SVM) then NB
and Decision Trees (DT).
The study in Reference [16] highlight the application of ML in industry for anomaly detection. The paper
compares various ML methods, including K-Nearest Neighbour (KNN), SVM, ANNs, and hybrid approaches.
Essentially, all the objectives are aimed at maximising efficiency and minimising costs, we will present an
example in Reference[34] of the use of an optimization model combing with ML to solve a manufacturing
scheduling problem with constraints of time and ressources. The model is formulated as a mixed-integer lin-
ear programming (MILP) problem aimed at minimizing the total duration (makespan) of production while
adhering to a set of intricate constraints. These constraints define conditions for the start, completion, and
interactions among operations, machines, workers, and configurations. The objective is to achieve an optimal
scheduling of production operations, accounting for Industry 4.0-specific constraints. Some assumptions are
considered to simplify the problem: No operation preemption, Single worker and machine focus, Full initial in-
formation. The authors are based on parameters such as: set of jobs, machines, operations and configuration...
In addition, authors in Reference [17] aim to introduce a prognostic method for predicting tool wear using
random forests (RFs). They assess the effectiveness of RFs when compared to feedforward backpropagation-
artificial neural networks (FFBP)-(ANNs) ,support vector regression (SVR) and in terms of performance for
this purpose. The comparison is based on empirical data gathered from conducting experiments. The results
indicate that RFs is better than FFBP-ANNs and SVR in terms of performance , showing precision in pre-
dictions for tool wear. The paper in [18] outline on a comprehensive approach of machine vision to automate
processes for detecting material category, to enhance the capabilities of technological tools, such as robots,
that are being used in manufucturing enviroment using method SVM. The paper in Reference [19] focuses on
the topic of the human-robot collaboration (HRC), where authors propose a robot model represented in an
interactive, dynamic interface using some deep learning algorithm of recognition and classification, including
three modules speech, movement and body recognition.

2.3 Major Application Areas of ML in industry 4.0

We present different application areas of Industrial ML: Optimizing Processes, Ensuring Quality Con-
trol, Predictive Maintenance , Human-Robot Collaboration and Ergonomics
These areas or topics were categorized, grouped and visualized in the mind map illustrated in Figure 2.
(a) Mind map of high potential value areas for ML in in-
dustry 4.0 (b) the targets of ML solutions in Industry 4.0.

Fig. 2: the targets of ML solutions in Industry 4.0.

3 Reinforcement learning

3.1 Reinforcement learning approch

Reinforcement Learning (RL) algorithms concerned two essential elements, an agent operating in a space. The
agent model interacts directly with the environment and manipulates it to perform a desired learning process.
According to [5], the , the agent start to receives the information about state called S of the environment.
Based on this state and its probabilistic decision process, the agent selects an action A. Subsequent to
executing the action in the previous state, the environment move to a novel state, and the agent is rewarded
R. This cycle of reinforcement learning persists, with the ultimate objective of maximizing, aiming to maximise
the rewards by improving the choice of actions towards an optimal control policy. In essence, the agent strives
to acquire the most effective control strategy for the environment.
RL holds great promise for achieving autonomous control in various industrial applications, such as
processes, machines and systems. In addition, multi-agent RL is gaining attention in the research community
and is showing potential for product control. Despite its potential, RL remains under-utilised in industry
compared to other machine learning techniques.

3.2 RL application in industry 4.0

This part provide a review of related work in the application of reinforcement learning in the areas of industry
4.0. RL has been applied effeciently in different areas in industry

Fig. 3: RL publications in industry 4.0 in the past five years

 Most articles included are recent publications (Figure3 ). The article in reference [21] propose a reinforce-
ment learning (RL) method for tracking use, concerning an agent trained to choose the best optimal state
based on an internal measurment to maximise the Rewards in each step. The applications of RL in industry
4.0 concerns often production scheduling and robotics. The study in Reference[22] propose an agent based on
Reinforcement learning trained to solve a dynamic scheduling problem (DJSP) . A comaprasion study of per-
formance for this proposed approach with other rules prove it efficiency and accuracy, depending on different
scenarios and criteria for this comparison. Q-learning[23] is frequently employed in various papers, including
the article in [24]. The article in Reference [25] shed light on scheduling optimization in manufacturing using
a plenty reinforcement learning algorithms based intelligent. In other context related to robotics in smart
manufacturing, some deepl RL methods such as Self-learning Processes are provided in [26], where authors
expirement a RL model named Deep Deterministic Policy Gradient (DDPG) for the control of time and
pressure injection. Moreover, authors in Reference [27] proposed a new neural network and Reinforcement
learning method known as Reinforcement Learning Unit Matching Recurrent Neural Network (RLUMRNN)
to solve a problem related to the performance control of neural network affected by the selection of hidden
layers. This part provide a review of previous works in the application of reinforcement learning in the areas
of industry 4.0. RL has have been used successfully in many applications.
An approch using deep reinforcement learning based on multiple agent, is proposed by authors in [28] to
solve Scheduling optimization model for energy system of a microgridpowered MEC network. Furthermore,
the use of RL for anomalie Dection has succesuful result presented in Reference [29], a framework combin-
ing RL and supervised learning, result a high intrusion prediction. In addition, RL is used in various other
applications. In class of prediction and forecasting , there is a predominance of method of planning relying
on value function approximation and decision processes of Markov. For detection policy gradient methods
are prominent. For evaluation, assessment, assignment, and resource management, multi-agent methods are
in the spotlight, signalling a shift towards distributing tasks to lower-level tools rather than centralised data
processes.
Then, in the initial category, the allocation of various RL methods closely resembles the general distribution,
while in the second category, policy gradient methods are predominant, as allocation-related problems tend
to favour the creation of optimal policies. In the Clustering, Decision Making, Classification, Scheduling, and
Planning classes, there is a contrasting pattern emerges: multiagent methods are underrepresented, suggest-
ing that research in these operations still tends towards centralised solutions. The paper in Reference [32]
tackles joint scheduling of tasks and optimizing cost. It explores machine learning vs. optimization methods,
proposing a stochastic dynamic programming approach and an attention-based deep reinforcement learning
model.
Figure 4 illustrates the core categories encapsulated by reinforcement learning methodologies.

Fig. 4: Distribution of RL methods by principal captured classes on [30]

4 CONCLUSION

It is widely acknowledged that AI and ML are assuming increasingly pivotal roles in the digital transforma-
tion of manufacturing towards Industry 4.0, contributing to enhanced quality, productivity, and operational
efficiency. In contemporary times, ML algorithms find extensive utility across diverse manufacturing domains,
including optimization, control and security. The heightened transparency achieved within the manufacturing
ecosystem through these technologies proves advantageous for cost reduction without compromising produc-
tion quality. RL methods also hold substantial promise in Industry 4.0 applications. One key rationale is that
new tools necessitate a level of optimization that surpasses what can be achieved through human interven-
tion. The present paper has furnished a comprehensive survey of the state-of-the-art ML and RL methods
developed since 2018. Recent trends reveal the ascendancy of RL-based approaches in the research landscape,
largely propelled by advancements in smart manufacturing. Various machine learning paradigms, including
supervised , unsupervised, deep learning, and reinforcement learning, are briefly introduced alongside their
applications within the manufacturing sector. To address these explorations, we undertook a literature re-
view, appraising the most recent scientific publications. As we delve into the potential of AI/ML and RL,
we remain optimistic about their capacity to drive sustainable growth, operational efficiency, and innovation
within the industries of the future.

References
1. M. McKnight, ”IoT, Industry 4.0, Industrial IoT: Why Connected Devices are the Future of Design,” 2017.
2. F. A. Germán, D. L. Santos, and A. N. Fabián, ”Industry 4.0 technologies: implementation patterns in manufac-
turing companies,” International Journal of Production Economics,Vol.210, pp 15-26, 2019.
3. Hernavs, J., Ficko, M., Berus, L., Rudolf, R., Klančnik, S., ”DEEP LEARNING IN INDUSTRY 4.0-Bref
overview”, vol.21, 2018.
4. J.Dalzochioa, R.Kunsta, E. Pignatonb, A.Binottoc, S.Sanyalc, J.Favillad, J.Barbosaa, ”Machine learning and
reasoning for predictive maintenance in Industry 4.0: Current status and challenges”, Vol.123, 2020
5. ChenT., Sampath V., May M.C., Shan S., Jorg O.J., AguilarMartı́n J.J., Stamer F., Fantoni G.,Tosello G., Calaon
M. ”Machine Learning in Manufacturing towards Industry 4.0: From ‘For Now’ to‘Four-Know’”. , vol 13, 2023
6. B. Bajic, I. Cosic, M. Lazarevic, N. Sremcev, A. Rikalovic, ”Machine Learning Techniques for Smart Manufac-
turing: Applications and Challenges in Industry 4.0”, pp 1-31, 2018
7. Iqbal H. Sarker, ”Machine Learning: Algorithms, Real World Applications and Research Directions”, vol.2, 2021
8. Jinjiang Wanga, Yulin Maa, Laibin Zhanga, Robert X. Gao b, Dazhong Wuc, ”Deep learning for smart manufac-
turing: Methods and applications”, Vol.48, pp 144-156, 2018
9. Francesco Zangaro, Stefan Minner, Daria Battini, ”A supervised machine learning approach for the optimisation
of the assembly line feeding mode selection”, International Journal of Production Research, vol.59,2020
10. Yanzhou Fu, Austin R.J. Downey, Lang Yuan, Tianyu Zhang, Avery Pratt, Yunusa Balogun, ”Machine learning
algorithms for defect detection in metal laser-based additive manufacturing”, Vol.75, pp 693-710, 2022
11. Huimin Zhao, Xiaoxu Yang, Baojie Chen, Huayue Chen, Wu Deng, ”Bearing fault diagnosis using transfer learning
and optimized deep belief network”,Measurement Science and Technology, vol.33, 2022
12. Abdul Wahid, John G. Breslin, Muhammad Ali Intizar, ”Prediction of Machine Failure in Industry 4.0: A Hybrid
CNN-LSTM Framework”, Appl. Sci., vol.12, 2022
13. A.Rivas, JM. Fraile, Pablo Chamoso, A.González-Briones, I.Sittón, JM. Corchado, ”A Predictive Maintenance
Model Using Recurrent Neural Networks”,advances in Intelligent Systems and Computing, vol.950, 2019
14. Dorina Weichert, Patrick Link, Anke Stoll, Stefan Ruping, Steffen Ihlenfeldt, Stefan Wrobel, ”A review of machine
learning for the optimization of production processes”, vol.104, pp 1889–1902, 2019
15. J.Schmitt, J.Bönig, T.Borggräfe, G.Beitinger, J.Deuse, ”Predictive model-based quality inspection using Machine
Learning and Edge Cloud Computing”, Advanced Engineering Informatics, vol.45,2020
16. Daniel Ramotsoela, Adnan Abu-Mahfouz, Gerhard Hancke, ”A Survey of Anomaly Detection in Industrial Wire-
less Sensor Networks with Critical Water System Infrastructure as a Case Study”, Sensors, vol.18, 2018
17. Dazhong Wu, Connor Jennings, Janis Terpenny, Robert X. Gao, Soundar Kumara, ”A Comparative Study on
Machine Learning Algorithms for Smart Manufacturing: Tool Wear Prediction Using Random Forests”, Vol.139
18. Durga Prasad Penumuru, Sreekumar Muthuswamy, Premkumar Karumbu, ”Identification and classification of
materials using machine vision and machine learning in the context of industry 4.0”, vol.31, pp 1229–1241, 2020
19. Hongyi Liu, Tongtong Fang, Tianyu Zhou, Yuquan Wang, Lihui Wang, ”Deep Learning-based Multimodal Control
Interface for Human-Robot Collaboration”, Vol.72, pp 3-8, 2018
20. Benjamin Rolf, Ilya Jackson, Marcel Müller, Sebastian Lang, Tobias Reggelin, Dmitry Ivanov, ”A review on
reinforcement learning algorithms and applications in supply chain management”, 2022
21. Yudha P. Pane, Subramanya P. Nageshrao, Jens Kober, Robert Babuška, ”Reinforcement learning based com-
pensation methods for robot manipulators”,Vol.78, pp 236-247, 2019
22. Ali Fırat ˙Inal, Ça ˘grı Sel, Adnan Aktepe, Ahmet Kür¸sad Türker, Süleyman Ersöz, ”A Multi-Agent Reinforce-
ment Learning Approach to the Dynamic Job Shop Scheduling Problem”, Sustainability, vol.15, 2023
23. Shiva Raj Pokhrel and Sahil Garg, ”Multipath Communication With Deep Q-Network For Industry 4.0 Automa-
tion and Orchestration”, Vol.17, pp 2852 - 2859, 2021
24. Bernd Waschneck, Andre Reichstaller, Lenz Belzner, Thomas Altenmuller, Thomas Bauernhansl, Alexander
Knapp, Andreas Kyek, ”Optimization of global production scheduling with deep reinforcement learning”, 51st
CIRP Conference on Manufacturing Systems, vol.72, pp 1264-1269, 2018
25. Ling Wang, Zixiao Pan, Jingjing Wang, ”A Review of Reinforcement Learning Based Intelligent Optimization for
Manufacturing Scheduling”, vol.1, pp 257-270, 2021
26. Rasmus E. Andersen, Steffen Madsen, Alexander B. K. Barloa, Sebastian B. Johansena, Morten Nøra, Simon
Bøgha, ”Self-learning Processes in Smart Factories: Deep Reinforcement Learning for Process Control of Robot
Brine Injection”,Vol.38, pp 171-177, 2019
27. F. Li, Y. Chen, J. Wang, X. Zhou, B. Tang, ”A reinforcement learning unit matching recurrent neural network
for the state trend prediction of rolling bearings”, Vol.145, pp 191-203, 2019
28. Md. Shirajum Munir, GSarder Fakhrul Abedin, Nguyen H. Tran,Eui-Nam Huh, and Choong Seon Hong, ”Risk-
Aware Energy Scheduling for Edge Computing With Microgrid: A Multi-Agent Deep Reinforcement Learning
Approach”, vol.18, pp 3476-3497, 2021
29. Xiangyu Ma, Wei Shi, ”AESMOTE: Adversarial Reinforcement Learning With SMOTE for Anomaly Detection”
,Vol.8, pp 943-956, 2021
30. Tamás Kegyes,Zoltán Süle,János Abonyi, ”The Applicability of Reinforcement Learning Methods in the Devel-
opment of Industry 4.0 Applications,” pp 1-31, 2021
31. Sabri Abderrazzak, Allaoui Hamid and Souissi Omar , ”Stochastic Dynamic Programming for Earliness-Tardiness
Single Machine Scheduling with Maintenance Considerations”, 2021
32. S.Abderrazzak, A.Hamid and S.Omar , ”Reinforcement learning and stochastic dynamic programming for jointly
scheduling jobs and preventive maintenance on a single machine to minimise earliness-tardiness”, 2023
33. Sabri Abderrazzak, Allaoui Hamid and Souissi Omar , ”Adaptive Large Neighborhood Search for the Just-In-Time
Job-shop Scheduling Problem”, 2022
34. Yuanyuan Li, Stefano Carabelli, Edoardo Fadda, Daniele Manerba, Roberto Tadei, Olivier Terzo1 ”Machine
learning and optimization for production rescheduling in Industry 4.0”, 2020
35. A. L. Samuel ”Some Studies in Machine Learning Using the Game of Checkers”, page 601-614, 1967
36. Mohammed M, Khan MB, Bashier Mohammed BE. ”Machine learning: algorithms and applications.” , 2016
37. Silver D, Huang A, Maddison CJ, Guez A, Sifre L, Van Den Driessche G, Schrittwieser J, Antonoglou I, Panneer-
shelvam V ”Mastering the game of go with deep neural networks and tree search. nature.”, 2016
38. Gao et al. ”Industry 4.0: A Survey on Technologies, Applications and Open Research Issues” , 2019
39. Massimo Bertolini a, Davide Mezzogori b, Mattia Neroni b, Francesco Zammori b ”Machine Learning for industrial
applications: A comprehensive literature review”, 2021