Dr. Jude W. Shavlik |
Department of Computer Sciences &
Department of Biostatistics & Medical Informatics,
University of Wisconsin-Madison
Jude W. Shavlik is a professor in the Department of Computer Sciences and the Department of Biostatistics & Medical Informatics at UW-Madison. He earned his Ph.D. at the University of Illinois in 1988 and joined the UW faculty that same year. His research interests include learning algorithms that take advice; relational learning; reinforcement learning; transfer learning; machine learning; and data mining applied to biomedical tasks such as disease diagnosis, protein-structure determination and information extraction from online biomedical text.
Thirty Years of Combining Symbolic and Numeric Learning
For nearly 30 years, Dr Jude's research group has investigated the use of domain knowledge, expressed in some version of mathematical logic, that is refined or exploited by numeric-based learning algorithms. These include what we called knowledge-based neural networks and knowledge-based support vector machines. Key ideas of these methods, as well as the behind-the-scenes motivations that lead to them will be covered in his session and also describes why we switched from using the phrase 'prior knowledge' to using 'advice.' Finally, some of their recent work on Markov Logic Networks, which can be viewed as a knowledge-based graphical model will be discussed.
Dr. Fabia U. Battistuzzi
Department of Biological Sciences
Dr. Fabia U. Battistuzzi is an Assistant Professor in Evolution at Oakland University. She obtained her PhD
from the Pennsylvania State University in 2007 in Evolutionary Biology and Astrobiology with a special
focus on bioinformatics. She has further specialized in evolutionary medicine during her postdoctoral
training. Her research interests include evolution of early life, evolution of diseases, phylogenomics, and
The role of Big Data in biology and medicine
The historical constraint on data acquisition that was characteristic of most biological fields has been lifted
by the advent of next generation sequencing technologies. The past decade has seen a dramatic increase
in data availability and types, from whole genome sequences to complex interactive maps of gene
functions and products, which have a high potential of revealing the inner workings of biological
organisms during their evolution, development, healthy, and diseased conditions. However, this vast
amount of information is often difficult to analyze and visualize and challenges are arising to identify
patterns within the data. A transdisciplinary approach to study biology is therefore key to improve our
power and accuracy to interpret data and reveal new dimensions to biological information. To achieve
this, it is fundamental to apply knowledge from computer science, graph theories, and data modelling to
the biological realm in order to interpret, test, and represent complex patterns in effective and accurate
ways. The growing fields of bioinformatics and evolutionary medicine are two examples of the
applications of Big Data science to biology to obtain patterns in organismal evolution, development,
genome function, and disease epidemiology. Discoveries within these fields are guiding major innovations
in space science, conservation biology, and medicine and are advancing our ability to use and interpret
genomes as data repositories of past events and predictors of future responses.
Dr. Mak Sharma
Head of School of Computing and Digital Technology
Birmingham City University
Head of School Mak Sharma is internationally recognised for his expertise in the use of vendor resources such as Cisco, Microsoft and Oracle, and embedding these into various qualifications. Mak joined the University over 20 years ago and during this period, he has worked on a number of projects the largest and most prestigious of which was the Millennium Point Project. Mak’s research interests are in the employability of graduates and his current technical interests are in the field of Network Management, and Health Informatics and Cloud.
IoT, Data and decision support
With more devices than the number humans on our planet (50B+ devices) being connected to Internet by 2020 and our every move captured by sensors of some kind,
whether that is in retail, healthcare or transportation, the captured data is of value to
someone. With the exponential increase in the use IoT devices, we need to harness the
collective power of the devices, the data and the resulting revenues that are potentially
huge. As with all new revolutions, there are positive benefits as well some negatives,
and we need engineers to manage the technical changes and business leaders to create
new business models for a better world.
In order to have a global impact, we need to start with multi disciplinary education,
showing the future entrepreneurs and engineers the value of IoT. We also need to
research the current use and future use of technologies to enhance society and the
world. The use of such disruptive technologies will give rise to undesirable outcomes
and we need to mitigate against these potential issues, such as the IP Camera hack in
the US recently that caused problems for some famous service providers.
So how are we helping the customer? there are well know examples such the Jonnie
Walker Blue bottle which has intelligence to detect it has been opened and hence
provide a personalised experience based on the use of a smart phone and the
dedicated message on each bottle. Other examples include the intelligent tabs on
Virgin cargo, or Rolls Royce adding more sensors to measure, monitor, maintain, the
engine parameters remotely for an optimised flight. In this last example machine are
communicating with machines, with little intervention from humans, this is because
we have set up the parameter limits to trigger other machine events and based on AI
we have implemented; the machines will make these decisions much faster, more
acute and more repeatable the we can. All this data and decision making can then
used to create more and better dedicated experiences.
In this keynote, the main discussion points are: how IoT is disrupting our current way
of life generally for the better; What IoT means from an academic perspective,
method of teaching using practical techniques based rather than theory and how to
encourage graduates to embrace IoT. Then summarisation of Decision Support
System, their uses and applications, and how these are used today and how we could
create more value from the data. Finally explanation about 2 research projects that fit
with the IoT domain and further examples of IoT applications that can exploited
Dr. Amit Kumar Saxena
Dept of Computer Science & Information Technology
Guru Ghasidas University
Prof Amit Kumar Saxena, was awarded Ph D in 1998 in Computer Science mainly on Artificial Neural Networks. He has several publications in National, International journals, IEEE transaction and proceedings of conferences. He has visited various countries for academic assignments such as Malaysia, Singapore, Kuwait, USA, Spain, Taiwan. Prof Saxena has delivered invited and keynote addresses at International Conferences in India and abroad. He has authored a book on C Programming; he is also reviewer and editor of journals and conferences. Presently Prof Saxena is a Professor and Head, Dept of Computer Science and Information Technology, G G Central University Bilaspur CG State of India. His research specialization includes Soft Computing, Computational Intelligence, Dimensionality reduction for feature selection, evolutionary computing.
Big Data, Cloud Computing and Internet of Things (IoT): Concepts
With the advent of ARPAnet, the viability of internet was brought to common people. By
and by the use of internet was extended to every area of society to transmit and receive
data. The internet served as the backbone to connect various computing devices mainly
desktop or laptop computers. With the tremendous development in communication
devices, now internet has become an indispensible icon to connect smart phones,
tablets, i-phones etc. Kevin Ashton in 1999 introduced a new term Internet of Things,
abbreviated as IoT in the context of Supply Chain Management. IoT is vigorously
expanded to its current form and mostly defined as an internet of all physical objects,
animate or inanimate. The connected objects can speak to each other via sensors, RFIDs,
without human intervention. Humans can control the physical objects, even the humans
(particularly patients, kids, physically challenged) for various purposes. The applications
of IoT have a variety of range.The individual homes; offices and electronic gadgets can
be monitored and controlled by IoT devices. Personal health care can also be watched
thru these devises. The elderly or small kids at homes in particular can be monitored by
sensors, actuators by doctors or even by their family members sitting in their respective
offices without need of engaging nurses to look after them. The smart environment is
another vital application of IoT.
Another significant development in the area of data based technology has been ‘big
data’. In fact big data and IoT are regarded as two faces of the same coin. The big data
refers to a situation where normal or conventional data management techniques such as
SQL cannot manage such data due to its huge size. It is well known how dependency on
software applications is increasing day to day by virtue of bio matrix devices, security
devices, biopsy of human beings, radar and other military based devices, share markets,
leisurely arrangements of people, travel reservation data, ERP etc. All such applications
will lead to heaping mammoth amount of data. It is assumed that about 50 billion
devices will be connected with internet through IoT, all these devices will also
accumulate huge amount of data. There will naturally be a threat of leaking private data
of individuals and organizations which will depend on these devices. The coordination
among these devices will need high attention. Moreover, how many connecting devices
one will wear or take care of.
With the tremendous software applications ready to enter in society, several support
systems will be required. Obviously industries will hire cloud services for their
requirements. Clouds not only provide secured storage but also services and platform
for smooth functioning by its users. A compromised server could significantly harm the
users as well as cloud providers. Unfortunately users may not be aware what type of
their information could be stolen, including credit card and social security numbers,
addresses, and personal messages.
Although technology on one end offers attractive and optimistic approaches to deal with
our routines in life yet on the other end also alarms us to be aware of its limitations to a
certain extent. Data however will be central in the operations of all such technological
developments. Other issues to be addressed while adopting big data, IoT and cloud
computing will be privacy, security, architecture, design and implementation,
networking, connectivity along with existing infrastructure.
Dr. S.D Madhu Kumar
Department of Computer Science and Engineering
Dr. S.D Madhu Kumar is an Associate Professor in the Department of Computer Science and Engineering at NIT Calicut. He obtained his PhD in Computer Science from IIT Bombay. His research specialization includes Distributed Computing, Cloud Computing, Big Data & Cloud Database and software Engineering.
Big Data Modelling and Querying with Graph Databases
Graphs are extremely useful in understanding a large variety of datasets and are widely used in the database world. Graph Databases use Graph structures like nodes, edges and properties to represent and store data of the real world. Edges represent relationships which directly relate data items in the data store. The relationships allow data in the store to be linked together directly and retrieved with a single operation unlike in the relational model , where links between data are stored in the data itself, and are obtained by searching for this data within the store and using operators like the Join.
Graph databases are designed to allow simple and rapid retrieval of complex hierarchical structures that are difficult to model in relational systems. Graph databases offer an extremely flexible data model. When compared to relational databases, where join-intensive query performance deteriorates as the size of the dataset increases, with a graph database performance tends to remain relatively constant, even as the size of the dataset grows bigger. This is because queries are localized to a portion of the graph. As a result, the execution time for each query is proportional only to the size of the part of the graph traversed to satisfy that query, rather than the size of the whole graph.
Relationships in a graph form paths. Querying—or traversing—the graph involves traversing through paths. Because of the fundamentally path-oriented nature of the data model, the majority of path-based graph database operations are highly aligned with the way in which the data is laid out, making them extremely efficient. With the growing importance of graph databases, there exists a variety of query languages for graph databases. GRAM, GraphDB, G, Cypher Query Language are few examples for Query languages used in this context.