Understanding the Internet of Things

Discuss about the Machine Learning To Make Intelligent Decisions In Internet Of Things.

The internet of things involve the connection and the interaction of a number of devices. A large number of devices are connected together in a network. All the devices are connected and the devices can interact with each other. The various applications of the internet of things such as the smart city or the smart street lighting system can easily be controlled by the use of machine learning. The machine learning concept compares the various data and puts forward the best possible solution. The concept of machine learning can also be used for finding the cyber security solutions. The report shows how the concept of machine learning technology can be used in internet of things for making intelligent decisions. The report also puts forward the methodologies

The internet of things concept is the interconnection of a large number of devices to a certain network that will help in the interaction with the other devices that are present in the network (Kim et al., 2016). The devices may involve the various sensors, actuators and the other important devices. The connection to the same network helps in interaction with the large number of devices at the same time. If any message is to be posted to a large number of systems at the same time then the use of the internet of things can be brought to practice. The machine learning process involves the use of the system in order to detect the possible outcomes from a particular project. The machine learning process is helpful in the execution of the various tasks. The new form of tasks that the machine may not be able to perform is recorded and the solution is told to the machine from the other sources. The solution, which was new to the machine is now stored in the machine such that if such a problem arises in the future then the task would be solved easily by the machine learning process. The machine learning may not be able to find some solutions or predict a number of things. The use of internet of things technology may be helpful in the finding of the various solutions. The system or the devices that are connected to the internet of things framework can be able to find the solution for the system.

The report discusses about how machine learning can be used in the decision- making intelligently in the internet of things network. The report presents a literature review, which puts forward the works in the topic, which is to be discussed. The report discusses the various methods that can be used in order to implement the machine learning process in the internet of things. The various methodologies are then compared on the various fields and then the best method that suits the project is then selected.

Machine Learning in Decision Making

Rapid developments in hardware, software as well as communication technologies have allowed the emergence of internet-connected sensory devices. It provides observation as well as data measurement from physical world. It is estimated that total number of internet-connected devices will be used between 20 to 25 billion by 2020 (Zhao & Ge, 2013). As the numbers increase and technologies become mature, the volume of date published will also increase. The internet-connected devices technology are referred as IOT that continues to extend current internet through providing connectivity as well as interaction between physical and cyber volumes (Zhang et al, 2014). Moreover, increased volume of IoT produces Big Data characterized by velocity in terms of location and time dependency.

Many of the organization cannot take decisions in a proper manner. Due to the wrong decisions taken by the organization many of the projects end up in failure. Thus the making of the proper decisions is important for the smooth running of the organization. The making of the intelligent decisions such as the use of the correct form of technology for the correct purpose is required. The main problem lies in the choice of the technology and the analysis of the technology. A proper cost analysis for the implementation of the technology should be done so that the organization does not have to face the problem in the later half.

The report discusses about the topic and the purpose that the report solves. The report discusses about the concept of the internet of things and the machine learning technology. The report then discusses how the machine learning process can be used for the taking of the intelligent decisions in the IoT in an organization. The literature review presents the use of the machine learning concept previously and presently for the making of the intelligent decisions. The use of the various organization has been made in order to properly understand the concept of implementation of the machine learning and show the works of the previous and the recent times. The three parts of the reports involve:

• The introduction of the report containing the purpose of the report, the structure of the report and the summary of the report
• The second part involves the literature review, which gives the various information on the basis of the past and the present works that is done.
• The last part of the report involves the various methodologies that are involved and the methodology, which is suitable for the present scenario.

The Machine learning process can be helpful in the identification of the process of the what will be good and what would affect in a bad manner. The use of machine leaning can be  made in order to know what decision has to be taken to benefit (Sridhar & Smys, 2017). The machine learning makes the use of various forms of techniques and helps in the identification of the future scope of the decision. The process of machine learning analyses a number of data at the particular topic before giving the best possible answers(Zhao & Ge, 2013). This is how the concept of machine learning can be made to make the important decisions in the field of internet of things. The internet of things can present with a number of solutions to a particular topic. Among these solutions the most efficient solution is to be found, which can be done with the help of machine learning process. The intelligent decision that is to be taken in this case is done with the help of machine learning. The machine learning process can help in the analyzing of the various data on the fields of the different solutions that are given. The data that best represents the positives for a particular solution is then selected. The mechanism of machine learning can be extended further into the various other fields(Ding et al.,2018). This will help in the analyzing of the data. The machine learning basically performs the task of the data analysis and puts forward the best possible solution based on the previous solution that had occurred in the past.

Literature Review

The Google makes the use of machine learning to continuously improve its status and position in the market. The Google provides the best search results and the most compressed search results by making the use of the machine learning technology. The machine learning technology implementation helps Google in providing the services it continues to provide over the years(Zhang et al, 2014). The concept of machine learning has also been applied to the various photos and the videos of the Google, which also helps in the management of the large number of photos and videos in an efficient manner. The machine learning procedure can also manage the huge amount of the data. The photos and the videos, which have to be shown to the are checked with the help of the machine learning. Thus, all these important and the intelligent decisions are taken with the help of the machine learning concept.

The project requires trained personnel for the carrying out the project as the technology has to be applied in a proper manner. The systems should be in a proper working condition as the improper functioning of one may lead to the improper functioning of the other systems. This may happen as all the systems are connected by the help of internet of technology. If any one of the system is vulnerable to attack then the other may also have a chance of getting attacked. All the requirements for the proper implementation of the machine learning concept should be

Iris is the other company, which makes the use of the machine learning in order to take the intelligent decisions that are involved in internet of things. The Iris company is responsible for the carrying out of the research and the development of various fields. The Iris organization is basically responsible for the finding out of those articles which are trying to suppress the importance of the relevant and the truthful source of the data that are present (Sadeghi, Wachsmann & Waidner, 2015). The company looks to the providing of the correct and the genuine information, which is hidden under the various articles that have been published with irrelevant information. There are articles, which are written by the various authors without the reference of the proper data. In this process the actual data gets lost and the wrong articles are reviewed. Iris is responsible for the holding on the actual data. People have been referring to Iris in order to get the relevant information. The company makes the working possible by following the machine learning process, which performs the survey on a number of data and calculates on an automatic manner. As the use of machine learning is made for the purpose of surveying the data it is possible for the organization to provide the correct information and the detailed survey of the data that are collected (Meidan et al, 2017). This also help the organization in getting hold of the market and there are a large number of customers who seek the help of the website in order to know about the various information and the data. Iris has even outclassed the Google Scholar.

Comparing Methodologies

The organizations by the implementation of the machine learning concept not only to gain the benefit in the market or to take important decisions in IoT but also in order to take the important decisions related to the internal works of the organization. By machine learning the ways in which the organization can be expanded and the profit of the business can be improved can also be done. The intelligent decisions are involved within the organization as well for the smooth running of the organization. This can be done by the method of the pattern recognition, which is a part of the machine learning(Sicari et al., 2015). By the help of the pattern recognition, one can check if everybody is doing the work in a proper manner or not. This can be done by the help of the tracking the work that is done.

The research methodology in first journal is nothing but a theoretical kind of defining and presenting the basics and various kinds of strategic decision from various kinds of resource in domain of field and science of business life(Dunaway et al,2017). However, this kind of approach contains certain number of advantages and disadvantages like those of effective practices.

From point view, various kinds of emphasis have been provided regarding subjective and individual kinds of findings (Zhu et al., 2015). The truth and reality does not relate to the fact of various figures and understanding considering various kinds of subjective areas which are generally encountered (Ding et al.,2018).

The actual idea or topic of Internet of things focus on standardization (Wu et al.,2014). Apart from that research is nothing but a combination of various kinds of research things. One of the major kinds of targets focus on qualitative kinds of research things, which can be used for generalization of findings in the form of conclusions. In this paper an idea has been provided regarding decision making for supply chain and things related to vice-versa.

In the second methodology the idea or concept of data and functionality, which can be accessed from any location and various kinds of device has been introduced (Bello &Zeadally, 2016). In this particular case the fitness bracelet generally acts like a IOT sensor along with providing means of accessing and consuming of data (Zhong et al., 2017). This particular device also subsumes other kinds of devices like pedometer by the help of software functionality. If the provided data is aggregated across a population of users and other kinds of datasets, new kinds of insights which can easily put focus on epidemiological kind of data, activity level of population, various kinds of lifestyles and data from demographic (Lee, Kao & Yang,2014). This kind of information has a strong kind of value from marketers, healthcare services, insurance firm or organization and government kind of agencies.

Case Studies of Machine Learning in IoT

Machine learning algorithm are generally used for making of prediction related to various kinds of data patterns (Posada, 2015). The same of machine learning and predictive algorithm are used for a basis for a number which is connected with intelligent type of consumer devices. Other kinds of consumer devices are generally leaned from various kinds of voice pattern. Optimization algorithm also makes use of various kinds of learning mechanism which is considered to be specific to both sensors and intelligent device which generally interact under various kinds of dynamic condition (Wang et al.,2016). This kind of parameters cannot go beyond certain number of parameters. The algorithm is responsible for sense, response and various kinds of adaptation.

In the methodology of thirdjournal the idea or concept of WSN. Machine learning is generally considered to be viable kind of solution for reducing capital value and operational kind of expenditure, as well as improving the life of network (Baron &Musolesi, 2017). Sensors are considered to be key things which helps in tracking of objects or things of the emerging kind of networks. The aggregation of various kinds of networks form a network whose ultimate goal is to generate and aggregation of data for various kinds of internal purpose. The data which is generated from the various kinds of sensors needs to be finely tuned prior to undertaking of any kind of analytical procedures. The majority of machine learning techniques makes use of combination kind of framework for feature extraction and modification of certain number of algorithms which are used for identifying handwriting or speech (Kumar, Goyal& Varma, 2017). There are various kinds of algorithm on sensory data like telemedicine, quality of air monitoring, localization on terms of indoor and smart kind of transportation (Liu & Jiang, 2016). Conventional machine learning has certain number of limitation like inability for optimization non-differential discontinuous loss functions along with feasible kind of training. 

Deep learning can be defined as a collection of various kinds of algorithm procedures with generally mimic the brain (Kim et al., 2017). This kind of layer generally provides hierarchical knowledge which is generally derived from simpler kind of knowledge’s. Real world problems are generally involved of processing of certain number multimedia of sensor data like speech or recognition of face which is considered to be challenging to digitally along with infinite number of problems (Sengupta et al., 2017). Problems are considered to be common kind of thing in processing of sensory data and are generally acquired from various multimedia

Challenges in Implementation

In the fourth journal we come across the purpose or idea of internet of things (IOT) which is used for development of smarter kind of environment along with simplified kind of lifestyle by saving time, energy and money (Jeschke et al., 2017). With the help of various kinds of technologies expenses in various industries are easily reduced. The enormous kind of investment and many kinds of studies runs along with IOT has made it top on the market in the recent years. IOT is generally considered to be set of some kinds of connected devices which can easily transfer data among one another can easily in optimization of performance. It also deals with various kinds of details related to human awareness or input. IOT of machine learning for intelligent decision making is mainly inclusive of four kinds of components like sensors, processing of networks, analyzing of data and monitoring of various components of system. The most important kind of advancement RFID tags were encountered when RFID (Radio frequency Identification) tags came into action. Lower cost sensors tend to become available, well developed along with changed communication protocol. IOT is generally integrated with various kinds of connectivity issues with various kinds of technologies and connectivity which is necessary (Sonntag et al., 2017). In IOT communication protocols are generally divided into three major kinds of components like

Device to Device: This kind of communication generally provides proper kinds of communication with various kinds of cellular networks.

Device to Server: In this particular kind of communication various kinds of data are sent to server, which is known to be far from various kinds of devices.

Server to Server:In this particular kind of communication server generally transmits various kinds of data between each other. This type of communication is generally applied to various kinds to various cellular networks.

In the methodology of last journal, the idea of NodeMcu is considered. It is found to be important for node because it will help in coding for various kinds of client’s servers so that data on various kinds of server can be sent easily. In the coming section an idea has been provided regarding Data Repository and Blynk Storage where data can be easily pushed to various kinds of cloud platforms. In this paper an idea has been provided regarding NodeMCUwhich is considered to be an open source of IOT based platform which consist of microcontroller. So it can operate independently and it can be easily programmed into various kinds of Arduino which is based on IDE. In the operating node NCU can easily provide SSID and various kinds of password related things.

Trained Personnel and Proper System Functioning

Comparative kinds of analysis have been done on the fact or idea regarding how many kind of nodes are connected. Various kinds of decision making algorithm is generally used for understanding or predicting the values. Based on level of ppm, humidity and other parameters like temperature this IOT device will decide regarding various kinds of parameters for improving of air quality and amount of trees which can be planted or right percent of greenery which needs to be increased in this particular kind of area.  In this journal idea has been provided regarding various kinds of sensors which can be used for detection of some specific kinds of things.

The goal of machine learning in the various kinds of programs generally focus on program computer which generally increases efficiency of experiences for solving a particular kind of problem (Karwowski, &Ahram, 2017). Various kinds of successful organization of machine learning makes use of experiences for solving a specific kind of problem. Various kinds of specific things related to machine learning has been used for various kinds of system which makes use of system for data for prediction of behavior of various customers. For a proper kind of treatment of machine learning various kinds of problems and solution has been provided. Various kinds of fields like statistics, recognition of pattern, neural networks and signal processing and data mining has been provided.

Some of most important kind of work which is being done is for reaping the value from Internet of things (IOT) which generally takes data into next level by making use of machine learning (Patel, Ali &Sheth, 2017). Similar to good kind of IOT based solution one can easily provide easy viewing of data. This ultimately relates building of algorithm which can be easily be predicted and it also addresses next kind of data which is used for historic failure of various kinds of data. Machines are generally considered to be good in various kinds of mechanical and physical process but it is considered to be a great kind of thing of various kinds of automated decision making in various kinds of complex environments.

Internet of things can be easily extended to distributed kind or processing of high volume of data (Oliveira, 2018). In every kinds of devices makes uses of sensors which can be used for data in repeated manner, challenges of big data can be easily encountered. For overcoming this kind of issues a distributive kind of networking can be easily used for dividing large number data into packets and after that each data is assigned to different kinds of processing (Bonnefoi et al., 2017). The distributed kind of network consist of two kinds of frameworks Hadoop and related spark. During migration from cloud to fog various kinds of issues can easily occur like:

• Reducing the load of network.
• Increasing the processing speed of various kinds of data.
• Reduction in the usage of CPU (Central Processing Unit)
• Reduction in the energy consumption.
• Processing of high volume of data.

Company Case Study: Iris

One of the biggest benefits which can be obtained from cloud computing is saving of cost in various kinds of IT organization. Various business irrespective of type or size general moved to cloud computing through reduction of various kinds of cost in various kinds of equipment (Li &Li , 2017). It generally provides additional kind of power over the internet which can easily replace machines of value of million dollars like servers. It reduces various kinds of parameters like hardware, licensing and renewal of fees and cutting of cost in terms of capital and operational cost. As a platform of cloud one can easily provide service for the thing which is needed the most. Cloud computing generally comes up with better kind of security which can easily be used for improvisation of protection of data.

There are large number of issues can be encountered like security, costing model, charging model and service level agreement and place of migration.

Security:It is generally cleared from the perspective that security issues played an important kind of role in the domain of cloud computing. Without any kind of doubt putting of data is considered to be an important kind of thing. It generally comes up with various kinds of issues like loss of data, phishing and lastly botnet.

Costing Model:Cloud based consumers may be considered to be important for computation and integration (Rathore et al., 2016). The cost of integration of data can be considered to be important part protocol and various kinds of interfaces.

Charging Model:From the idea or logic of various kinds of elastic resources final kind of idea has been provided regarding regular kind of data centers. For Sass based cloud providers the cost involved with various kinds of offering. This is mainly inclusive of redesign and redevelopment.

Service level Agreement:Various kinds of cloud consumers does not have any kind of control on availability, resources and performances of this kinds of things.

Feasibility

Cloud computing things can be easily used for redefining of data and various kinds of uses related to it (Fadlullah et al., 2017). Sharing of file can be considered to be any easy fact which can be easily used for making various kinds of things much easier. If someone needs a file, then one can easy grab it from any location. It is only possible from a list of devices like mobile phone and tablets. Before making of proper kinds of decision one can easily make use of large number of benefits from various kinds of benefits of cloud services. Various kinds of small business round the globe can easily make use of IT engagement by properly working some kind of load.

Pattern Recognition in IoT for Organizational Decision Making

Connectivity

Emergences of consumer based IOT things like security is considered to be an important parameter. A lack in any one of the above kind of parameter may led to certain number of nodes and low dependent requirements. It ultimately leads to increasing number of complexity in each kind of nodes. It also follows security related issues in each kind of node. Various kinds of new challenges are encountered during securing of data with various kinds of software defining networks. There are certain number of physical separation which is in control with various virtual environments which are over the multiple network verses the traditional kinds of networks. Other kinds of challenges are encountered in securing multiple channels for various kinds of communication in comparison to traditional kind of networks (Zhong et al., 2016). The various kinds of challenges associated with employment and distribution of wealth in ever increasing domain of automation. Another kind of issues which can be encountered is the reliability of various internet connection. The last kind of issue encountered is machine learning is the security of various kinds of threats associated with it. This kind of data generally require increasing of sophisticated data analytic and various kinds of algorithm which can be used for learning.

The methodology that can be used in order to implement the technology of the machine learning in internet of things for the making of the efficient and the intelligent decisions is cloud computing. Cloud computing can be used as a strong methodology that can be used by the machine learning technology for the making of the intelligent and smart decisions. There are various data and the activities, which are stored in the cloud and is retrieved from the cloud(Pop, 2016). With the help of the machine learning technology all the previous data can be retrieved and all the previous files may be can be edited. Machine learning is responsible for the storage of the large number of information and also comparing all these previous forms of data so as to provide the best solution possible. The solution is chosen on the basis of the best quality data that is available. Now, for instance the cloud, which is a very efficient form of storage of information and the data has the information that can lead to the solution of the problem. All the data has to be checked and all those data have to be compared so as to reach the solution. For this comparison the use of the machine learning technology has to be made. With the machine learning technology the large chunk of data that is present in the cloud can be analyzed and then can be processed in order to obtain the best solution that is available. The clod computing is not only responsible for the storage of the data but it is also responsible from the cyber security purpose. Thus, in case of the occurrence of any cyber attack or any other issue in the system in the organization, then again by the help of the machine learning technology the data about the previous solutions to the problem can be retrieved. The cloud would be holding the solutions to the various problems of cyber attack and thus the record would also be present in the cloud on how to eliminate the cyber attack(Sridhar & Smys, 2017). Thus, by the help of the machine learning technology all the solutions that are present to eliminate the particular problem that has come up would be searched and then the best possible solution in order to eliminate that problem is put forward. Thus, it is evident that use of the machine learning can be made in the field of internet of things as the cloud computing is a part of the internet of things technology. All the devices in an organization are also connected with the medium of cloud. The machine learning also holds a major part in the making of the intelligent decisions(Porambage et al, 2016).  As in this case the machine learning holds an important space in not only the comparison of the data about the information but also for providing the best solution as to get rid of the risk. The cloud computing is an important part of every organization nowadays and forms an integral methodology of the internet of things.

Conclusion

The cloud is also responsible for the holding the value of the various private and the public keys. These keys are most important part of the cloud computing technology. The cloud computing technology connects with the various systems by the help of the method of these private and the public keys(Terry, 2016). The private keys are given to the people who are integral part of any of the organization or are an important part of the business. The confidential contents of an organization is only accessible to the people who have the private keys numbers. If these numbers of the private keys get into the wrong hands then the data of the organization has to be compromised. Thus, the use of machine learning is made in order to keep track of the data of the activities of the users of the keys. This will help in accessing as to who is trying to access the confidential files in an improper way. The activities may be tracked and if any obscene activity is taking place then the IP address of the machine can also be tracked thus helping the prevention of the loss of important data.

The quantitative data research methodology is used by the organization in order to make the intelligent decisions. The machine learning is used in this case as well for making the analysis of the data. The machine learning process is responsible for the analyzing of the all forms of data and then gives the result of the problem. The quantitative form of the methodology helps in the complete analyzing of the all forms of data that are available. As by the method of internet of things all the systems are connected all the data that are present in all the computers of the organization or all the information that were entered into the system is checked compared and then the result is put forward(Baron & Musolesi, 2017). The quantitative analysis involves the use of the maximum number of data for providing the result. If the maximum number of sources of data are used then the result would be a better one. A more number of comparison would be involved in the quantitative form of data thus providing a much better result of the analysis.

The other form of data research technique is the qualitative data research, which involve the research of the best quality data. The number of data researched is not important in this case but the quality of the data matters(Sarker et al, 2015). However, in this case the number of the data sources researched matters and the quality of the data does not matter. The quality of the data does not matter in this case as the data are already checked before and have been earlier applied. The searching from the more number of sources is important rather than the looking for quality. The reason is that the machine learning technology will be able to compare from the large number of data sources. All these data sources will contain the solutions that were previously used and have been successful, thusthe analysis has to be done from the data sources and the best solution is put forward.

Conclusion:

From the report it can be concluded that the machine learning in IoT can be used to make the various important decisions. These decisions not only involve the decision on expanding the business or how to make more profit but also involves the other important decisions as well. The other decisions may involve are the prevention of the breach of the data by the external or the internal user. The retrieval of the data and the comparison of the data is the important part for the processing of the data. The internet of things comprises of a large and complex connection of the systems. Thus, the systems are responsible for the storage of large number of information and the data cannot be easily analyzed and cannot be easily checked. Thus, for the checking and the analysis of the data the concept of the machine learning technology is used in order to analyze and compare the large chunk of data easily and fast. The best solution is put forward in an automatic manner by using the method of machine learning. The various methodologies are implemented, which helps in the making the intelligent decisions in an easy manner. The implementation of the cloud computing technology has been put forward as this methodology can use the machine learning in the IoT to make the important decisions. The methodology can be used not only to analyze the large amount of data but also to provide the solution of the risk, which may occur to the organization. The machine learning is thus helpful for providing the solutions to the various decision on the basis of the data analysis.

References :

Baldini, G., Peirce, T., Botterman, M., Talacchini, M. C., Pereira, A., & Handte, M. (2015). Iot governance, privacy and security issues. Position paper, European Research Cluster on the Internet of Things.

Baron, B., &Musolesi, M. (2017). Interpretable Machine Learning for Privacy-Preserving IoT and Pervasive Systems. arXiv preprint arXiv:1710.08464.

Bello, O., &Zeadally, S. (2016). Intelligent device-to-device communication in the internet of things. IEEE Systems Journal, 10(3), 1172-1182.

Bonnefoi, R., Besson, L., Moy, C., Kaufmann, E., &Palicot, J. (2017). Multi-Armed Bandit Learning in IoT Networks: Learning helps even in non-stationary settings.

Ding, G., Wu, Q., Zhang, L., Lin, Y., Tsiftsis, T. A., & Yao, Y. D. (2018). An amateur drone surveillance system based on the cognitive Internet of Things. IEEE Communications Magazine, 56(1), 29-35.

Dunaway, M., Murphy, R., Venkatasubramanian, N., Palen, L., &Lopresti, D. (2017). Research Agenda in Intelligent Infrastructure to Enhance Disaster Management, Community Resilience and Public Safety. arXiv preprint arXiv:1705.01985.

Fadlullah, Z. M., Tang, F., Mao, B., Kato, N., Akashi, O., Inoue, T., &Mizutani, K. (2017). State-of-the-art deep learning: Evolving machine intelligence toward tomorrow’s intelligent network traffic control systems. IEEE Communications Surveys & Tutorials, 19(4), 2432-2455.

Jeschke, S., Brecher, C., Meisen, T., Özdemir, D., &Eschert, T. (2017). Industrial internet of things and cyber manufacturing systems. In Industrial Internet of Things (pp. 3-19). Springer, Cham.

Karwowski, W., &Ahram, T. (Eds.). (2017). Intelligent Human Systems Integration: Proceedings of the 1st International Conference on Intelligent Human Systems Integration (IHSI 2018): Integrating People and Intelligent Systems, January 7-9, 2018, Dubai, United Arab Emirates (Vol. 722). Springer.

Kim, H. J., Chang, H. S., Suh, J. J., & Shon, T. S. (2016). A study on device security in IoT convergence. In Industrial Engineering, Management Science and Application (ICIMSA), 2016 International Conference on (pp. 1-4). IEEE.

Kim, R. G., Doppa, J. R., Pande, P. P., Marculescu, D., &Marculescu, R. (2017). Machine Learning and Manycore Systems Design: A Serendipitous Symbiosis. arXiv preprint arXiv:1712.00076.

Kumar, A., Goyal, S., & Varma, M. (2017, July). Resource-efficient Machine Learning in 2 KB RAM for the Internet of Things. In International Conference on Machine Learning (pp. 1935-1944).

Lee, J., Kao, H. A., & Yang, S. (2014). Service innovation and smart analytics for industry 4.0 and big data environment. Procedia Cirp, 16, 3-8.

Li, B., & Li, Y. (2017). Internet of things drives supply chain innovation: A research framework. International Journal of Organizational Innovation (Online), 9(3), 71B.

Liu, C., & Jiang, P. (2016). A cyber-physical system architecture in shop floor for intelligent manufacturing. Procedia Cirp, 56, 372-377.

Meidan, Y., Bohadana, M., Shabtai, A., Guarnizo, J. D., Ochoa, M., Tippenhauer, N. O., & Elovici, Y. (2017). ProfilIoT: a machine learning approach for IoT device identification based on network traffic analysis. In Proceedings of the Symposium on Applied Computing (pp. 506-509). ACM.

Oliveira, E. (2018). Beneficial AI: the next battlefield. Journal of Innovation Management, 5(4), 6-17.

Patel, P., Ali, M. I., &Sheth, A. (2017). On Using the Intelligent Edge for IoT Analytics. IEEE Intelligent Systems, 32(5), 64-69.

Pop, D. (2016). Machine learning and cloud computing: Survey of distributed and saas solutions. arXiv preprint arXiv:1603.08767.

Porambage, P., Ylianttila, M., Schmitt, C., Kumar, P., Gurtov, A., & Vasilakos, A. V. (2016). The quest for privacy in the internet of things. IEEE Cloud Computing, 3(2), 36-45.

Posada, J., Toro, C., Barandiaran, I., Oyarzun, D., Stricker, D., de Amicis, R., … &Vallarino, I. (2015). Visual computing as a key enabling technology for industrie 4.0 and industrial internet. IEEE computer graphics and applications, 35(2), 26-40.

Rathore, M. M., Ahmad, A., Paul, A., & Rho, S. (2016). Urban planning and building smart cities based on the internet of things using big data analytics. Computer Networks, 101, 63-80.

Sadeghi, A. R., Wachsmann, C., & Waidner, M. (2015). Security and privacy challenges in industrial internet of things. In Design Automation Conference (DAC), 2015 52nd ACM/EDAC/IEEE (pp. 1-6). IEEE.

Sarker, A., Ginn, R., Nikfarjam, A., O’Connor, K., Smith, K., Jayaraman, S., & Gonzalez, G. (2015). Utilizing social media data for pharmacovigilance: a review. Journal of biomedical informatics, 54, 202-212.

Sengupta, U., Doll, C. N., Gasparatos, A., Iossifova, D., Angeloudis, P., Baptista, M. D. S., … & Luo, J. (2017). Sustainable Smart Cities: Applying Complexity Science to Achieve Urban Sustainability.

Sicari, S., Rizzardi, A., Grieco, L. A., & Coen-Porisini, A. (2015). Security, privacy and trust in Internet of Things: The road ahead. Computer Networks, 76, 146-164.

Sonntag, D., Zillner, S., van der Smagt, P., &Lörincz, A. (2017). Overview of the CPS for smart factories project: deep learning, knowledge acquisition, anomaly detection and intelligent user interfaces. In Industrial Internet of Things (pp. 487-504). Springer, Cham.

Sridhar, S., & Smys, S. (2017, January). Intelligent security framework for iot devices cryptography based end-to-end security architecture. In Inventive Systems and Control (ICISC), 2017 International Conference on (pp. 1-5). IEEE.

Terry, D. (2016). Toward a new approach to IoT fault tolerance. Computer, 49(8), 80-83.

Wang, H., Xu, Z., Fujita, H., & Liu, S. (2016). Towards felicitous decision making: An overview on challenges and trends of Big Data. Information Sciences, 367, 747-765.

Wu, Q., Ding, G., Xu, Y., Feng, S., Du, Z., Wang, J., & Long, K. (2014). Cognitive internet of things: a new paradigm beyond connection. IEEE Internet of Things Journal, 1(2), 129-143.

Zhang, Z. K., Cho, M. C. Y., Wang, C. W., Hsu, C. W., Chen, C. K., & Shieh, S. (2014). IoT security: ongoing challenges and research opportunities. In Service-Oriented Computing and Applications (SOCA), 2014 IEEE 7th International Conference on (pp. 230-234). IEEE.

Zhao, K., & Ge, L. (2013). A survey on the internet of things security. In Computational Intelligence and Security (CIS), 2013 9th International Conference on (pp. 663-667). IEEE.

Zhong, R. Y., Lan, S., Xu, C., Dai, Q., & Huang, G. Q. (2016). Visualization of RFID-enabled shopfloor logistics Big Data in Cloud Manufacturing. The International Journal of Advanced Manufacturing Technology, 84(1-4), 5-16.

Zhong, R. Y., Xu, C., Chen, C., & Huang, G. Q. (2017). Big Data Analytics for Physical Internet-based intelligent manufacturing shop floors. International Journal of Production Research, 55(9), 2610-2621.

Zhu, N., Diethe, T., Camplani, M., Tao, L., Burrows, A., Twomey, N., … & Craddock, I. (2015). Bridging e-health and the internet of things: The sphere project. IEEE Intelligent Systems, 30(4), 39-46.