Decision Processes in Smart Learning
Environments
Peter Mikulecky
(
B
)
University of Hradec Kralove, 50003 Hradec Kralove, Czech Republic
peter.mikulecky@uhk.cz
Abstract.
Smart learning environments can be naturally considered to
be a new challenging step of computer enhanced learning evolution, offer-
ing many new interesting facilities. Smart environments in general as
well as their special case smart learning environments can be studied as
being based on a sophisticated multi-agent based architecture, which is
behind all the decision processes that appear in the environment and
enable the environment’s functionality. The aim of this paper is to sum-
marize recent state of the art in the area, with strong focus on decision
processes in smart learning environments. We are not presenting any new
algorithms for these processes, but we discuss several various possibili-
ties and approaches instead. The paper intends to be a starting point
in looking for new ways or approaches for creating decision processes by
which smart learning environments cope with the problem of multiple
residents in a smart (learning) environment.
Keywords:
On-line learning
·
Smart environments
·
Multi-agent archi-
tectures
·
Decision processes
1
Introduction
We are witnessing now a big effort focused on research in the area of technological
support of education. Most recent step in this effort resulted in the concept
of smart learning environments, illustrated by the number of interesting and
really implemented projects. Smart learning environments, based on approaches
and technologies deployed in related areas of Ambient Intelligence and Smart
Environments, certainly deserve considerable attention of the large community
oriented on technology enhanced learning. Smart learning environments can be
naturally considered to be a new challenging step of computer enhanced learning
evolution, offering many new interesting facilities. Moreover, a number of new
information technologies and approaches certainly will influence research in the
area of smart learning environments. Let us mention cloud based architectures
and applications as a representative.
As [
1
] pointed out, besides its technological perspective, social perspective,
or ethical perspective, it is possible to study the area of Ambient Intelligence
also from an educational perspective. This educational perspective deals with
problems and challenges related to proper education in relevant areas.
c
Springer International Publishing Switzerland 2016
N.T. Nguyen et al. (Eds.): ICCCI 2016, Part II, LNAI 9876, pp. 364–373, 2016.
DOI: 10.1007/978-3-319-45246-3 35
Decision Processes in Smart Learning Environments
365
The famous ISTAG Report [
2
] started in 2001 a decade of various research
initiatives in the rapidly growing area of ambient intelligence. Among the four
basic scenarios described in the report, it introduced also a smart environment
example in the form of the
Scenario 4: Annette and Solomon in the Ambient
for Social Learning
. That was a vision of a learning environment, based on a
position that learning is a social process.
According to the original description in the ISTAG report [
2
],
the Ambient
for Social Learning (ASL) is an environment that supports and upgrades the
roles of all the actors in the learning process, starting with the roles of the men-
tor and the students as most concerned parties. The systems that make up the
ASL are capable of creating challenging and interacting learning situations that
are co-designed by the mentor and students in real-time. Students are impor-
tant producers of learning material and create input for the learning ‘situations’
of others. In other words, the ASL is both an environment for generating new
knowledge for learning and a ‘place’ for learning about learning.
Of course, the Ambient for Social Learning should be also a physical space
(consisting of a room or even several rooms in a building) together with all of
its ambient facilities, including many linkages with similar places. Its layout and
furnishing has to be flexible and diverse, so that it can serve the learning purposes
of many different kinds of groups and individuals. Such an Ambient for Social
Learning could be an intelligent classroom supporting and upgrading roles of all
the actors in the learning process, with a special accent on the roles of the mentor
and students as the most important roles. The Ambient for Social Learning is
conceived as a ‘learning system’ that is growing and improving simply by using
it. Nevertheless, discussing the importance of the
ISTAG Annette and Solomon
scenario
, Alsaif et al. [
3
] pointed out, that if we took into account this ISTAG
scenario that could serve as an ideal case for a smart learning environment,
the popular view of
anywhere and anytime learning
should be considered as
impractically broad. We already discussed that in [
4
], where an agent-based
architecture for the original ASL was proposed.
An attempt to specify smart learning environments in a slightly detailed way
was published in [
5
]. A bit more general overview of the Ambient Intelligence
possibilities in education brings our recently published paper [
6
]. The aim of
the paper was to identify and analyze key aspects and possibilities of Ambient
Intelligence applications in educational processes and institutions (universities),
as well as to present a couple of possible visions for these applications. The con-
clusion of the presented research was that exploitation of Ambient Intelligence
approaches and technologies in educational institutions was possible and could
bring us new experiences utilizable in further development of various Ambient
Intelligence applications.
In the scope of a recent research project we intend to go deeper into the nature
and structure of decision processes conducted by autonomous multi-agent archi-
tectures that are behind intelligent environments of various kinds (e.g., intelli-
gent offices, intelligent workplaces, smart learning environments, etc.), enabling
their proper functionality. Structures of these decision processes are being
366
P. Mikulecky
investigated and new knowledge about their nature should be obtained. In this
paper decision processes of autonomous multi-agent architectures of smart learn-
ing environments are studied, with a special focus on their specific problems.
2
Related Works
2.1
Smart Environments for Learning
According to [
7
],
a learning environment can be considered smart when the
learner is supported through the use of adaptive and innovative technologies from
childhood all the way through formal education, and continued during work and
adult life where non-formal and informal learning approaches become primary
means for learning.
That is, Kinshuk supports the meaning of smart learning
environments as neither pure technology-based systems nor a particular peda-
gogical approach, but a subtle mixture of both.
In [
8
],
a smart environment for learning is defined as any space where ubiq-
uitous technology informs the learning process in an unobtrusive, social or col-
laborative manner.
Alsaif et al. [
3
] pointed out, that a smart environment can
be an
aware
room or building, capable of understanding something about the
context of its inhabitants or workers; it can be a digitally enhanced outdoor
space park, cityscape or rural environment; or it can be the environment created
when peoples meetings or interactions are augmented by wearable devices.
Another interesting opinion can be found in [
9
]. They understand Smart
Learning Environments as
systems that apply novel approaches and methods on
the levels of learning design and instruction, learning management and organi-
zation, and technology to create a context for learning that provides learners with
opportunities for individualized learning and reflection in a motivating way, and
that allow teachers to facilitate learning, providing scaffolding and inspiration
based on the learners needs and a careful observation of her learning activities.
As a conclusion, they pointed out that approaches in the direction of Smart
Learning Environments cannot be restricted only on the technological level, but
should involve also another dimensions.
According to [
10
],
a smart learning environment not only enables learners
to access digital resources and interact with learning systems in any place and
at any time, but also actively provides the necessary learning guidance, hints,
supportive tools or learning suggestions to them in the right place, at the right
time and in the right form
. The features just mentioned should be essential
for smart learning environment. A smart learning system can be perceived as a
technology-enhanced learning system that is capable of advising learners to learn
in the real world with access to the digital world resources. A smart learning
environment aims to help students gaining knowledge even when they are doing
leisure activities. Hwang concludes that a simple incorporation of an intelligent
tutoring system into a context-aware ubiquitous learning environment is not
enough for obtaining a real smart learning environment.
Decision Processes in Smart Learning Environments
367
2.2
Decision-Making in Smart Learning Environments
Kinshuk and his colleagues [
11
] stressed, that there are three major features of
the development of smart learning environments that separates smart learning
environments from other advances in learning technologies. The three features
(or directions) are:
– full context awareness,
– big data and learning analytics,
– autonomous decision-making.
Another important feature of smart learning environments, which differ them
from other learning environments, is their autonomous knowledge management
capability that enables them to automatically collect individual learners life
learning profiles. Collected learning profiles can track the learning progress of
each individual learner over quite long periods and across the range of sources
of evidence about the learners progress. According to Kinshuk et al. [
11
], based
on individual life learning profiles and the techniques of big data and learning
analytics, smart learning environments can precisely and autonomously analyze
learners learning behaviors in order to decide in real time, for example,
– what interactions with the physical environment to recommend to the indi-
vidual learners to undertake various learning activities, as well as the best
location for those activities,
– which problems the learners should solve at any given moment,
– which online and physical learning objects are the most appropriate,
– which tasks are the best aligned with the individual learners cognitive and
meta-cognitive abilities,
– what group composition will be the most effective for each group members
learning process,
– and so on.
Such autonomous decision making and dynamic adaptivity has according to
Kinshuk the potential to generalize and infer learners learning needs in order
to provide them with suitable learning conditions. This capability of decision
making and the capability of being dynamically adaptive to the learners’ needs,
based on learning analytics utilization, seems to be essential for a really successful
smart learning environment.
As Spector [
12
] points out, it is necessary for a smart learning environment
to autonomously provide
different learning situations and circumstances, as... a
human teacher or tutor... to help learners become more organized and aware of
their own learning goals, processes and outcomes
.
2.3
Decision-Making in Smart Environments
Following the concept of ubiquity, Ambient Intelligence is focused on technologies
and approaches for development of intelligent environments aiming at supporting
368
P. Mikulecky
their users, that is, persons surrounded by these environments. Decision processes
in intelligent environments are, for the sake of simplicity, usually studied in
specific types of environments, like smart homes or smart offices.
Architectures of these environments are usually modeled by multi-agent sys-
tems, facilitating thus investigation of processes enabling functionality of the
environments. Therefore, in general it is possible to study decision processes
in smart environments as decision processes provided by the underlying multi-
agent system, that are usually focused on fulfilling requirements or needs of
users surrounded by the environment. The situation is relatively easy in the case
of a single user, as it is no need to solve different, frequently also contradic-
tory requirements of individuals, situated in the environment. The single user
case can be taken into account in such situations, like, e.g., a senior is living in
her smart household, or a smart hospital room is monitoring the only patient
recovering herself in the room.
Of course, there are some interesting exceptions also in the case of one human
situated in a smart room. As Becerra and Kremer presented in [
13
], also several
other types of inhabitants could be taken into account, especially non-human,
e.g., animals, plants, but even valuable paintings, or furniture requiring some
specific treatment. If these non-human inhabitants are considered to be nearly
equally important as their human house-mate, it would be necessary to take also
their requirements into account and include them into decision processes of the
respective smart environment. An example introduced in [
13
] could illustrate
the case:
When the human inhabitant is not in the room, the environmental
conditions in the room (e.g. light, temperature and humidity) are not monitored.
This could have adverse effects on the other (non-human) inhabitants of the
room. A dark room could prevent the plants from growing; a cold room would
make the leather sofa uncomfortable to sit on when the human inhabitant returns;
a warm, bright room could damage the valuable paintings.
Just described case of one human and several non-human inhabitants of a
smart environment can certainly be solved as a case of single resident, with
several co-residents having just a restricted number of relatively standardized
requirements. It could be clear from the fact, that those other inhabitants are
important but are not able to interact with the environment. Their possible
requirements could be in a sense inserted into the smart environment decision
making facility. For instance, a plant do not like bright light and temperature
below 20 C. The smart environment, when monitoring the daylight and temper-
ature, could always take these restrictions into account.
The multiple resident case in intelligent environments, in contrary to the
single resident case, is very interesting and far more difficult. As Cook and Das
[
14
] pointed out, a lot of mobility tracking algorithms worked well just for single
resident case. The multiple resident case still has been not solved satisfactorily.
It was proven already that optimal location tracking of multiple residents is a
NP-hard problem [
15
]. Nevertheless, as Cook and Das [
14
] argued further on,
it can be supposed that each resident in an intelligent environment behaves
selfishly in order to fulfill her/his own preferences or objectives and to maximize
Decision Processes in Smart Learning Environments
369
her/his utility. Therefore, the appearance and activities of multiple residents in
an intelligent environment might lead to conflicting goals and serious problems
if not deficiencies in decision-making behavior of the whole intelligent system.
An intelligent environment must be intelligent enough in order to cope with such
kind of problems by striking a balance between multiple preferences, in many
cases having even contradictory nature. In [
15
], the problem of location tracking
of multiple residents was investigated from the perspective of stochastic game
theory; however, this solution seems to be just a theoretic result still without
practical application.
3
Decision Processes
3.1
Introductory Remarks
Decision processes as such are studied broadly with a focus on their nature and
usage in various organizations, see, e.g., [
16
] or [
17
], usually as Markov decision
processes. Decision processes in multi-agent architectures are studied as well,
see, e.g., [
18
,
19
], or [
20
], usually with a focus on a particular application. Quite
often approaches, as multi-agent reinforcement learning, are used [
21
]. There is
also possible to use modeling and simulations of multi-agent systems for studying
structure and actual impact of particular decision processes, see, e.g., [
22
].
Agent-based simulation is now a well established simulation modeling tool
in academia and on the way to achieving the same recognition in industry, as
Siebers and Aickelin [
22
] pointed out. Agent-based simulation is well suited to
modeling systems with heterogeneous, autonomous and pro-active actors, such
as human-centered systems. A special position among them have smart environ-
ments, which certainly are human-centered, their actors (agents) are autonomous
up to a very high level and their important task is to be pro-active in serving
the users residing in such an environment.
Luck et al. [
23
] made a distinction between two
Multi-Agent System
par-
adigms:
multi-agent decision systems
and
multi-agent simulation systems
. In
multi-agent decision systems, agents participating in the system must make joint
decisions as a group. Mechanisms for joint decision-making can be based on
economic mechanisms, such as an auction, or alternative mechanisms, such as
argumentation. Multi-agent simulation systems are used as models to simulate
real-world domains where agent-based modeling is appropriate.
3.2
Decisions of Smart Learning Environments
First, let us go back to the ISTAG
Scenario 4: Annette and Solomon in the
Ambient for Social Learning
. According to the ISTAG group, a number of spe-
cific technologies would be needed for implementation of this Smart Learning
Environment, among others the following ones:
– Recognition (tracing and identification) of individuals, groups and objects.
370
P. Mikulecky
– Interactive commitment aids for negotiating targets and challenges (goal syn-
chronization).
– Natural language and speech interfaces and dialogue modeling.
– Projection facilities for light and soundfields (visualization, virtual reality and
holographic representation), including perception based technologies such as
psycho-acoustics.
– Tangible/tactile and sensorial interfacing (including direct brain interfaces).
– Reflexive learning systems (adaptable, customisable) to build aids for review-
ing experiences.
– Content design facilities, simulation and visualization aids.
– Knowledge management tools to build community memory.
An important problem in each smart environment is the problem of how the envi-
ronment evaluates users needs and how it assigns preferences to them. Actually,
when many users are involved in a ubiquitous environment, the decisions of one
user can be affected by the desires of others. This makes learning and predic-
tion of user preference difficult. To address the issue, Hasan et al. [
24
] propose
an approach of user preference learning which can be used widely in context-
aware systems. The approach based on Bayesian RN-Metanetwork, a multilevel
Bayesian network to model user preference and priority is used there.
According to [
24
], a real smart system should have the following three capa-
bilities:
1. A smart system should be able to do inference.
2. A smart system should be able to learn by itself. User and developers can act
like teachers, but the knowledge should be improved incrementally.
3. A smart system should be able to solve some difficult problems, such as the
conflict among the users.
When applying the first capability on the case of smart learning environments,
the environment should monitor the users (in this case the learners making use of
the environment), and on the basis of collected data and using inferences from the
data the smart learning environment should decide about such matters as what
learning activities could be recommended to individual learners, or what are the
problems the learner should solve in the given moment. Certainly this capability
is useful in the case when assessing learners’ skill should be provided by the
smart learning environment itself [
25
]. In this case decisions of the environment
could influence heavily the further study path of each individual learner.
If we think about the second capability, certainly it is expected that a really
smart learning environment has to learn important facts related to the learn-
ers’ progress by itself. If this is possible, then the environment would be able to
decide, e.g., what group composition will be the most effective for each group
members learning process (cf. [
11
]). Such a learning of the smart learning envi-
ronment could be achieved by using various modern knowledge management
approaches, see, e.g., [
1
,
26
], or [
27
].
The third capability seems to be most important from the research behind
this paper point of view. Solving difficult problems, as conflict among the users
Decision Processes in Smart Learning Environments
371
of the smart learning environment, is certainly a very difficult task. These con-
flicts usually are based on users’ preferences that are changing over the time.
It is clear that always there is certain set of preferences of the particular user,
which are nearly static, not changing in a long term period. For instance, if a
person does not like to strong light, it is unlikely, that a day after this preference
will change. Such preferences can be evaluated by the environment quickly and
possible conflicts with others could be solved simply by comparing the prefer-
ences and using some of simple comparative algorithms (the biggest one is the
winner, or something like that). However, usually this is not the case of a group
of learners residing in a smart learning environments and having very often con-
tradictory requirements as to use some resources, or collaborating mutually on a
project, which can be a source of numerous disputes and disagreements. There it
is still lack of methods for solving such conflicts, and new approaches should be
found and applied. According to [
24
],
the preference of user changes over time
or based on situation. It makes online learning (or adaptation) a crucial require-
ment. Sometimes there is uncertainty in users temporary preference. User does
not always select the most weighted choice. Again, when there are many users in
the smart environment, the action of one user can affect others choice. It raises
the challenges of distinguishing the preference of each user as well as resolving
the conflicts among different user preferences. The introduction of probabilistic
model can handle these uncertainty and adaptive prioritization of users.
This
adaptive prioritization of users seems to be a promising way for handling the
users’ preferences successfully, however, this is still not solved satisfactorily.
4
Conclusions
Studying decision processes in smart learning environments can be one of ways
how to understand their functionality and adjust the underlying multi-agent
based architecture accordingly. It seems, that one of the most important points
here lies in investigating various methods how the users’ preferences can be
represented, evaluated, and compared mutually aiming to give the smart envi-
ronment a possibility of as proper decisions about its future actions as it is
possible. Probabilistic approaches, as the one by [
24
] could be promising, how-
ever, it is necessary to elaborate them further on. There could be some promising
directions based on ontologies, see, e.g., [
28
], nevertheless, the referenced paper
was not written with that aim. So, a lot of further research seems to be ahead.
The acknowledged project DEPIES is one of such initiatives aiming to con-
tribute to optimization and better coordination of decision processes oriented on
smart environment activities focused on the multiple residents case. And smart
learning environments are typically used by a group of residents, so they are a
good target for focused research with the described orientation. In the project,
new approaches and algorithms for decision processes in suitable multi-agent
architectures are investigated using the multi-agent modeling and simulation of
these architectures. In order to enable proper investigation of various decision
processes in a typical multi-agent based architecture of an intelligent environ-
ment, the modeling and simulation environment AnyLogic [
29
] is used broadly.
372
P. Mikulecky
Some most recent results in this direction have been published already in [
30
].
We believe, that agent-based simulation will be a strong tool in our way towards
deeper understanding of decision processes in smart environments, with a special
accent on smart learning environments.
Acknowledgment.
The support of Czech Science Foundation GA ˇ
CR #15-11724S
DEPIES is gratefully acknowledged.
References
1. Bureˇs, V., ˇ
Cech, P., Mls, K.: Educational possibilities in the development of the
ambient intelligence concept. Problems Educ. 21st Century
13
, 25–31 (2009). ISSN
1822-7864
2. Ducatel, K., Bogdanowicz, M., Scapolo, F., Leijten, J., Burgelman, J.C.: Scenar-
ios for ambient intelligence 2010, ISTAG report, European commission. Institute
for Prospective Technological Studies, Seville, November 2001.
ftp://ftp.cordis.lu/
pub/ist/docs/istagscenarios2010.pdf
3. Alsaif, F., et al.: Determination of smart system model characteristics for learning
process. In: International Conference on Convergence Technology, vol. 4, pp. 881–
885 (2014)
4. Mikuleck´
y, P.: Smart learning environments - a multi-agent architecture proposal.
In: 10th International Scientific Conference on Distance Learning in Applied Infor-
matics (DIVAI 2014), Wolters Kluwer (2014)
5. Mikuleck´
y, P.: Smart environments for smart learning. In: DIVAI 2012 (2012)
6. Bureˇs, V., Tuˇcn´ık, P., Mikuleck´
y, P., Mls, K., Blecha, P.: Application of ambient
intelligence in educational institutions: visions and architectures. Int. J. Ambient
Comput. Intell. (IJACI)
7
(1), 94–120 (2016)
7. Kinshuk: Roadmap for adaptive and personalized learning in ubiquitous environ-
ments. In: Kinshuk, Huang, R. (eds.) Ubiquitous Learning Environments and Tech-
nologies. Lecture Notes in Educational Technology, pp. 1–13. Springer, Heidelberg
(2015)
8. Winters, N., Walker, K., Rousos, D.: Facilitating learning in an intelligent envi-
ronment. In: The IEE International Workshop on Intelligent Environments, pp.
74–79. Institute of Electrical Engineers, London (2005)
9. Libbrecht, P., M¨
uller, W., Rebholz, S.: Smart learner support through semi-
automatic feedback. In: Chang, M., Li, Y. (eds.) Smart Learning Environments.
Lecture Notes in Educational Technology, pp. 129–157. Springer, Heidelberg (2015)
10. Hwang, G.J.: Definition, framework and research issues of smart learning
environments-a context-aware ubiquitous learning perspective. Smart Learn. Env-
iron.
1
(1), 1–14 (2014)
11. Kinshuk, Chen, N.S., Cheng, I.L., Chew, S.W.: Evolution is not enough: revolu-
tionizing current learning environments to smart learning environments. Int. J.
Artif. Intell. Educ.
26
(2), 561–581 (2016)
12. Spector, J.M.: Conceptualizing the emerging field of smart learning environments.
Smart Learn. Environ.
1
(1), 1–10 (2014)
13. Becerra, G., Kremer, R.: Ambient intelligent environments and environmental deci-
sions via agent-based systems. J. Ambient Intell. Human. Comput.
2
(3), 185–200
(2011)
Decision Processes in Smart Learning Environments
373
14. Cook, D.J., Das, S.K.: How smart are our environments? An updated look at the
state of the art. Pervasive Mob. Comput.
3
(2), 53–73 (2007)
15. Das, S.K., Roy, N., Roy, A.: Context-aware resource management in multi-
inhabitant smart homes: a framework based on nash H-learning. Pervasive Mob.
Comput.
2
(4), 372–404 (2006)
16. Pettigrew, A.M.: The Politics of Organizational Decision-Making. Routledge, New
York (2014)
17. Sigaud, O., Buffet, O.: Markov Decision Processes in Artificial Intelligence. Wiley,
New York (2013)
18. Yang, X., Yao, J.: Modelling multi-agent three-way decisions with decision-
theoretic rough sets. Fundam. Inf.
115
(2–3), 157–171 (2012)
19. Burnett, C., Norman, C., Sycara, K.: Trust decision-making in multi-agent sys-
tems. In: Proceedings of the 22nd International Joint Conference on Artificial
Intelligence, IJCAI 2011, pp. 115–120 (2011)
20. Yu, C.H., Werfel, J., Nagpal, R.: Collective decision-making in multi-agent sys-
tems by implicit leadership. In: Proceedings of the 9th International Conference
on Autonomous Agents and Multiagent Systems, vol. 3, pp. 1189–1196. Interna-
tional Foundation for Autonomous Agents and Multiagent Systems (2010)
21. Wu, J., Xu, X., Zhang, P., Liu, C.: A novel multi-agent reinforcement learning
approach for job scheduling in grid computing. Future Gener. Comput. Syst.
27
(5),
430–439 (2011)
22. Siebers, P.O., Aickelin, U.: Introduction to multi-agent simulation. arXiv preprint
(2008).
http://arxiv.org/abs/0803.3905
23. Luck, M., McBurney, P., Shehory, O., Willmott, S.: Agent technology: computing
as interaction (a roadmap for agent based computing). Technical report, University
of Southampton, UK (2005)
24. Hasan, M.K., Anh, K., Mehedy, L., Lee, Y.-K., Lee, S.-Y.: Conflict resolution and
preference learning in ubiquitous environment. In: Huang, D.-S., Li, K., Irwin, G.W.
(eds.) ICIC 2006. LNCS (LNAI), vol. 4114, pp. 355–366. Springer, Heidelberg (2006)
25. Klimova, B.: Assessment in smart learning environment – a case study app-
roach. In: Uskov, V.L., Howlett, R.J., Jain, L.C. (eds.) Smart Education and
Smart e-Learning. Smart Innovation, Systems and Technologies, vol. 41, pp. 15–24.
Springer, Heidelberg (2015)
26. Mikulecky, P.: User adaptivity in smart workplaces. In: Pan, J.-S., Chen, S.-M.,
Nguyen, N.T. (eds.) ACIIDS 2012, Part II. LNCS, vol. 7197, pp. 401–410. Springer,
Heidelberg (2012)
27. Mikuleck´
y, P.: Learning in smart environments - from here to there. In: 10th Euro-
pean Conference on e-Learning, pp. 479–484. ACL, Reading (2011)
28. Uskov, V.L., Bakken, J.P., Pandey, A.: The ontology of next generation smart
classrooms. In: Uskov, V.L., Howlett, R.J., Jain, L.C. (eds.) Smart Education and
Smart e-Learning. Smart Innovation, Systems and Technologies, vol. 41, pp. 3–14.
Springer, Heidelberg (2015)
29. Tuˇcn´ık, P., Bureˇs, V.: Inclusion of complexity: modelling enterprise business envi-
ronment by means of agent based simulation. Int. Rev. Model. Simul. (IREMOS)
6
(5), 1709–1717 (2013)
30. Mis, K., Cimler, R., Mikulecky, P.: Agent-based simulation for identifying the
key advantages of intelligent environments for inhabitants with special needs. In:
Sulaiman, H.A., Othman, M.A., Othman, M.F.I., Rahim, Y.A., Pee, N.C. (eds.)
Advanced Computer and Communication Engineering Technology: Proceedings of
ICOCOE 2015. Lecture Notes in Electrical Engineering, vol. 362, pp. 1031–1041.
Springer, Heidelberg (2016)
Document Outline
Do'stlaringiz bilan baham: |