Кумеков С.Е., Алинов М.Ш.,
Казахский национальный технический университет им. К.И. Сатпаева,
г. Алматы, Республика Казахстан
СТРАТЕГИЯ КАЗАХСТАН – 2050: ИНЖЕНЕРНЫЕ КАДРЫ ТЕХНОЛОГИЧЕСКИХ
УКЛАДОВ БУДУЩЕГО
Рассматриваются научные аспекты Стратегии Казахстан – 2050, в виде соответствия
Кондратьевским циклам роста, а также волнам технологических укладов, оказывающих влияние на
долгосрочное развитие экономики. Анализируются современные тенденции образовательных и
исследовательских программ в соответствии с принципами устойчивого развития, «зеленой»
экономики, высоких и энергоэффективных технологий зарубежных университетов. Обосновывается
необходимость перевода казахстанских образовательных программ по подготовке инженерных
кадров в соответствии с Болонским процессом на зарубежные аналоги и модели.
Будущая модель образования XXI века включает в себя два «опережающих фактора». 1.
Опережающее развитие самого образования (ориентированного на цели устойчивого развития) по
сравнению с другими сферами деятельности (экономической, политической и др.). 2. Опережающий
механизм в самом образовательном процессе, его ориентация на будущее и формирование модели
«зеленого устойчивого общества»[4]. В зарубежных странах в системе высшего образования
практикуются такие образовательные программы: устойчивое развитие и «зеленая» экономика,
окружающая среда и здоровье человека, «зеленая» экономика и органическое сельское хозяйство,
«зеленая» энергия, «зеленый» бизнес, экотуризм, «зеленая» химия и экология, материалы для
энергоэффективности и энергосбережения, энергоэффективная архитектура, экология и природные
ресурсы, возобновляемые источники энергии, изменение климата и территориальное развитие и др.
Исходя из этого, необходим пересмотр всех учебных программ, планов, специальностей,
государственных образовательных стандартов и других материалов под углом зрения проблем
будущего; особое внимание должно уделяться идеям устойчивого развития, управления природными
ресурсами, «зеленой» экономики, устойчивой энергетики, внедрения в производство высоких
технологий.
В Казахстане, как и в других странах СНГ в системе высшего образования готовятся кадры по
традиционным специальностям «Экология», «Безопасность жизнедеятельности и защита
окружающей среды», «Землепользование», «Геология», «Водные ресурсы и водопользование»,
«Биология», «Недропользование», «Электроэнергетика» и др., которые базируются на устаревших
ресурсо- и энергоемких стандартах. Необходим поворот к подготовке специалистов нового
технологического уклада и «зеленой» экономики с расширением доступности массового образования.
Научный потенциал страны будет существенно расширяться за счет модернизации и
переориентации университетской науки. Именно на приоритеты «зеленой» экономики и высоких
технологий важно направить усилия научного потенциала. В Казахстане медленно, но происходит
поворот в сторону увеличения финансирования НИОКР на «зеленые» проекты. Из более 7 млрд.
тенге грантового государственного финансирования более четверти, так или иначе, связаны с
новыми энергоэффективными технологиями. Проведение EXPO-2017 «Энергия будущего» следует
считать фактором основательной переориентации Казахстана на ценности «зеленого» развития. По
образцу Назарбаев университета будут созданы несколько университетских центров являющихся
кластерами инновационных технологий, среди них Казахский национальный технический
университет им. К.И.Сатпаева и Казахский национальный университет имени аль-Фараби. Для
существенного увеличения результатов научных исследований и их коммерциализации планируется
довести уровень финансирования науки до 3% к ВВП страны.
167
Секция 2
Новые информационные технологии (облачные вычисления, мобильные и мультимедийные
технологии, технологии по распознаванию речи, информационной безопасности и другие
технологии): отечественный и мировой опыт.
UCD 519
Zari Dzalilov
Federation University , Victoria, Australia
Email: z.dzalilov@federation.edu.au
INFORMATION TECHNOLOGY FOR LEARNING ENVIRONMENTS AND
COMMUNICATION STRATEGIES DESIGN IN CONNECTION TO BIG DATA
Abstract. The importance of interdisciplinary and collaborative learning and research has long been recognised
as a valuable strategy for learning. The opportunities for global networking have become much more accessible through
relatively recent advances in technology and transport.
We live in a transnational community where engagement with colleagues across continents is a regular
occurrence. However, interdisciplinary and inter cultural collaborations often present challenges in terms of finding a
common set of understandings and language to progress discovery and innovation. This paper discusses the importance
ofInformation Technology for designing learning environments and drawing upon multiple modes of communication
strategies to ensure that diverse viewpoints and perspectives are captured and understood.
Introduction
1. The importance of interdisciplinary and collaborative learning and research has long been
recognised as a valuable strategy for learning. We propose to consider the following research issues for the
platform, to build the skeleton of the paper : computational methods and data sets on Cystic Fibrosis,
Tobacco Control system, and Brain Complex Networks
The strategies are drawn from experiences working on two major transnational health studies where
complex mathematical processes of “optimization” and “data mining” are being applied to large datasets.
The paper also illustrates how the strategies devised to design these optimal learning environments align with
the recent neurological findings on how the brain functions and how learning is best facilitated. This is the
most interesting problem, and it is still a very new area of research despite of too many researchers
concerned with the problem of the function of the brain. Various models designed for understanding ofthis
unique element of the human body, that controls the rest of the functions of body systems.
Learning environments in connection to big data .
21 century can be called as the Era of Big Data. The explosion of available data in a wide range of
application domains creates a new challenges and opportunities in disciplines – ranging from science and
engineering to biology and business. The major challenge is how to take advantage of the unprecedented scale
of data, in order to learn the secrets of nature, and based on this knowledge to design learning environments for
a joint research. Going through much iteration within all aspects of the research framework we can get more
precise results for the problem presented by the data. The noise of the data is another important issue, and
sometimes we need eliminating the data to get the true pattern of the data to shine through
Big data in the biological and biomedical sciences: bio data mining challenges and opportunities.
Big data, by definition, challenges our ability to move, store, manage, retrieve and analyse information to
maximize knowledge generation. These informatics, information technology and data analysis challenges are
compounded by the complexity of problems that are of practical importance. Our future success in bio data
mining will depend critically on our ability to embrace and confront both the scale and the complexity of
bigdata.The setting for this paper originates in research linked to two major health studies:
A. Cystic Fibrosis : Australian data base –the first project , following up to 26 countries data set with
33000 patients’ records-European Cystic Fibrosis Patient registry data base
B. Tobacco smoking habits: 4 countries data set: Australia, UK, Canada and USA
The story oftwo ongoing projects related to Tobacco Control Systems and Cystic Fibrosis is a good
example of the learning processes. We use it as illustration fora study case of learning environment design,
168
based on a joint research of experts in different fields, such as medical doctors, mathematicians, information
technology experts, psychologists, data base managers and administrative staff for decision making. These
two projects based on different models, because of countries involved, main languages, and some other
cultural factors. This is also the process of learning environment design for two different large scale data sets
on Cystic Fibrosis and Tobacco Control Systems.In both cases large amount of data were available and the
challenge for the disciplinary teams was to make sense of the DATA.
A. SOME KNOWLEDGE ABOUT CF
Cystic fibrosis is the most common fatal genetic disorder in the Caucasian population. This is a terrible
disease with the life expectancy not over 46 years, mostly spent in hospitals for undergoing treatment.
Prognosis and diagnosis of CF are two important global issues for this health problem. Clinical scoring
systems for the assessment of Cystic fibrosis disease severity have been used for almost 50 years without
being adapted to the milder phenotype of the disease in the 21st century. A fresh approach is needed for the
development of comprehensive CF disease severity scales, which may be used as a disease predictor. The
main goal in the research on CF is to develop a scoring system to assess the longitudinal process of Cystic
Fibrosis. The team of mathematicians from Australian Universities and doctors from RCHproposedto
develop a new clinical scoring system by employing various computational methods: statistical, data mining,
and optimisation methods.
Despite significant developments in this area there is still a lot of evaluation work to be performed due
to the fact that medical data sets are diverse and it is difficult to formulate a unique criterion for all of them.
Optimization plays a fundamental role in designing efficient data mining techniques.
1. Data sets and design of testing environments, based on different approaches.
We previously identified an approach for developing a disease severity scale. That evaluation was
based on the Cystic Fibrosis database from the cohort at the Royal Children’s Hospital (RCH) in Melbourne.
The methods applied to this data set were the Linear Least Squares Fit (LLSF) and the Heuristic Algorithm
for Feature Selection.They allow analysing data sets with an arbitrary number of classes. However, the data
set from RCH was small, and more data points we needed to finalize a clinical score, by re-running these
methods in the larger data set. We had to design some learning techniques for applications that give us better
results in comparison with existing methods.
We proposed to refine this scale by using a hybrid model combining mathematical optimisation and
data mining approach. The advantage of these methods is that they allow one to consider datasets with an
arbitrary number of classes. Comparison of computational results of different methods was effective for
evaluation of mathematical optimisation methods for the solution of feature selection problems.
2. Joint research is a chance for the team members to learn from each other.
It was a grant based research, for the team, completed by researchers from UoB, and doctors from
RCH. The team members have been excited about joint research that was a chance for everybody to learn
from each other: mathematicians from doctors and vice versa. Doctor Gaudenz Hafen is currently the
Director of Paediatric Pulmonology Unit CF Centre in Switzerland, Lausanne. In terms of time, it was only
one year small grant for this research, but because of our interest, and passion of the doctors involved, we did
not give up. We obtained some preliminary reasonable results, which can be used by clinicians. It was only
first step towards the main goal, but we have proved that our methods work well!
Fantastic goal: creating bridges from global theories to applications, was a chance for us to obtain
some preliminary but very promising results , which can be extended then to clinical research, and from
clinical trials there should be a VERY POSITIVE WAY to THERAPY…..
Knowledge discovery
However, small data sets are not efficient; we had to access large scale data sets. Following the advice
of Dr. Gaudi, we applied for the access to USA data set on CF. We included also one team from USA for
this application, that was the mistake, and we failed with our application, although we learned from the
referees reports about “the strength of the team”, and “the methods applied”. One more drop to a further
success of the strong team! Different projects took over, but few years later on, using different methods to
the same data, we have got better results, that I presented at the Conference in Venice, where I was invited
for the round table discussions for the session “Social Mechanisms for Better Information Discovery and
Interpretation”
Teleconference organised by Committee of ESCFPR:Granted access to Data
After this conference, and further discussions with Gaudi, we decided to apply for the access to
European Society CF Patient Registry ( ESCFPR) database. This application was met very friendly by the
Committee of ESCFPR . We had the Teleconference to correct the details of the application. The
169
Teleconference was also a chance for us to learn not only from the experts in the field, but from
administrative staff as well. Few months later on we have been informed about the success of the application
that gives us the access to the ESCFPR database from 26 countries with 33000 records! This data will be
transferred to Australia, and I am only one, who has the access to this data base
Funds for research: Recent grant application to German fund
The access to to the ESCFPR database was again only the first step for a new wave of research on
“Real big data sets” on CF. The current situation at Australian Universities unfortunately doesn’t
support the projects with no funds. Only financially supported projects are very welcomed. Finally we found
the German Fund “Mukoviszidose”, that is supporting research projects on CF. On 13 December we
submitted the first round application for the project: “Clinical scoring systems for the assessment of Cystic
fibrosis disease severity”. We passed the first round successfully! 87% of all applications were cut down, but
we are given a chance to take a part at the final stage in this process. We had to submit more detailed
application to the fund by April. We have been informed that one more fund from Switzerland is interested
in this project that sounds great!In April I visited the Children Hospital in Lausanne for discussions with
doctors about details of the project.
B. Tobacco Control Systems
Background.Recent statistics on tobacco and health reveal that about 1.l billion people currently
smoke cigarettes, 80% of which lives in low and middle-income countries. Overall, the latest global statistics
show that a third of the male adult population smokers and smoking-related diseases kill one in ten adults,
which translate into five million premature deaths per annum; if current trends continue, smoking will be
responsible for one in six deaths by 2030. Development of theoretical and methodological frameworks in
data analysis is fundamental for modelling complex tobacco control systems.
Global picture. In response, significant progress in tobacco control policy planning and development
has been reported, especially in developed countries. At the moment, almost every jurisdiction in the world
has to join to the tobacco control battles and enormous efforts, like policy interventions, mass media
campaigns and the provision of smoking cessation information, have been made to cope with that problem.
The Framework Convention on Tobacco Control (FCTC)established by the World Health
Organisation in 2003 was the first international treaty devoted to public health. Up to now 142parties, which
represent 95% of the world’s population, have ratified the FCTC.
Tobacco Control data set
The International Tobacco Control Policy Evaluation Survey (ITCPES), [ITC survey, 2010] is a recent
coordinated international research and evaluation effort. This project provides massive survey data collected
from many countries including Australia, for studying and evaluating the psychosocial and behavioural
impact of diverse tobacco control policies to smoker behaviour across these countries (Figure 1).
Figure1 – A structure of the data set under consideration
170
The Framework Convention on Tobacco Control (FCTC) has been ratified by up to 142parties [].
Many countries have incorporated FCTC policies into their laws. These countries have attempted to
influence the behaviour of smokers by regulating and implementing diverse tobacco control policies. [].
In this study, we were interested to find clusters: groups of smokers with similar demographics,
responses to anti-smoking advertisements and warning labels, beliefs about quitting in the role of predicting
the rate of quitting attempt, etc. We aim to understand better the psychosocial and behavioural impact of
diverse tobacco control policies to distinct clusters of these smokers. Controlling tobacco smoking and
determining effective policies is difficult because of the complexity of human nature and behaviours. Also,
the success of tobacco control is not the result of single policies, but is the outcome of interactions among
various policies in various domains. Therefore, cluster analysis is helpful for dealing with causalities among
a set of stable clusters defined by fixed number of instances over the set of variables. In order to analyse this
data set we apply modified global k-means clustering algorithm to a survey data sets about a complex
tobacco control system.
Methodology development
Subsequently, developing most suitable methodologies and techniques to monitor the performance and
evaluatethe effectiveness of relevant tobacco control policies have becomeimportant research issues, since an
efficient and effective monitoring and evaluating system can provide accurateand timely information on the
performance of policies, programs and projects. This information can provide invaluable support for
decision-making, decision-refinement and ongoing management of government activities, and can underpin
accountability relationships. In order to describe the non-linear relationships more effectively, new global
optimization-based approaches were previously proposed in the paper [Z. Dzalilov et al., 2010]. Our
preliminary results indicate a possibility for a global optimal approach to covering all possible solutions in a
complex tobacco control system. (see [Z. Dzalilov et al., 2010] for more details).
Research
As the evidence of importance relating to the research topic above, considerable tobacco experts,
practitioners and academic researchers around the world have been involved in comprehensive tobacco
control research. There have also been several recent efforts to coordinate international research and
evaluation effects, like Global Youth Tobacco Survey and International Tobacco Control Policy Evaluation
Survey (ITCPES).
Methods: Data Mining and Optimisation
We applied optimisation based data mining techniques developed at CIAO research Centre toData
from the International Tobacco Control Policy Evaluation Survey (ITCPES).
The purpose of this project is to evaluate the psychosocial and behavioural impact of key national-
level tobacco control policies over the years. However, if we take a closer look at these research results, we
find that these approaches have fundamental limitations, since the outcomes presented in these papers did not
provide any mechanism to deal with complex non-linear data, which characterise most aspects of the tobacco
domain. In addition, these available techniques did not address the causal interrelatedness of smokers, non-
smokers, researchers, doctors, advocates, tobacco industries, policy makers etc in co-producing the targets of
tobacco control policies. As a result, significant gaps in tobacco control policy planning and development
remain.
Global networking, team work
(VicHealth Centre of Tobacco Control , The Cancer Council Victoria , and the team of researchers
from the Centre for Informatics and Applied Optimisation (CIAO), University of Ballarat (UB). This project
is the best example for illustration useful links between research, policy development and the effects of those
policies to the systems under consideration (in particular, in tobacco control systems). This project is about
unique combination of theoretical frameworks, survey data and expert judgments to develop innovative
models of the relationship between research knowledge, current policies and subsequent outcomes for
smoke-free policies, aimed at reducing the risk from environmental tobacco smoke.
Достарыңызбен бөлісу: |