MSLDS: Making Sense from Large-scale Data Streams
Evolving for almost a decade, big data is no longer just about managing and processing terabytes or even petabytes of data. Instead recent research and applications of large-scale analytics focus more on accelerating and improving the decision making process and more specifically making sense from large-scale, multivariate, complex data streams. As standard machine and statistical on-line learning methods cannot efficiently handle the complex structure and size of massive data, fulfil computational cost-effective requirements, large-scale, on-line analytics receive a growing interest both from academia and commercial companies.
- Background and Goals
- Topics of Interests
- Program Committees
- Instructions of Authors
- Important Dates
- Paper Submission
- Contact Us
Motivation: Evolving for almost a decade, big data is no longer just about managing and processing terabytes or even petabytes of data. Instead recent research and applications of large-scale analytics focus more on accelerating and improving the decision making process and more specifically making sense from large-scale, multivariate, complex data streams. As standard machine and statistical on-line learning methods cannot efficiently handle the complex structure and size of massive data, fulfil computational cost-effective requirements, large-scale, on-line analytics receive a growing interest both from academia and commercial companies. Data sets rapidly increase their size as they are often generated in a form of incoming multivariate streams. Sensor networks, environmental monitoring, traffic management, telecommunication, web log analysis, and anomaly/novel detection in enterprise network perimeters are examples of such applications where machines working in dynamic environments continuously generate data.
Challenge: Compared to static environments, the processing of data streams implies new requirements for on-line and incremental learning algorithms based on one scan of incoming instances. An important aspect of learning from streams is the ability to deal with changes in the data distributions and target concepts over time and in discerning these from random noise. Detecting these changes quickly and adapting the learning models to concept drifts becomes one of the challenges for learning algorithms over streaming data. Furthermore learning models from massive and stream data intersects with other related problems such as detecting rare cases, outliers, semi-supervised learning, dealing with missing data, non-available or partially available streams, statistical data compression, learning over dependent data, and transfer learning.
Aim: This special issue aims to share new research in defining and showcasing the value of large-scale stream mining, on-line and progressive analytics. We solicit research papers presenting new ideas, methods, and algorithms for large-scale on-line stream mining algorithms and applications.
- Online and incremental machine learning and large-scale stream mining
- Detection and adaptation to concept drift in complex data streams
- Discovery, detection and classification of complex patterns in massive or evolving data
- Scaling-up learning algorithms
- Near real-time analysis of massive data
- Applications to real-life problems from medicine, bioinformatics, multimedia, sensors, social networks and related domains
- Christos Anagnostopoulos, University of Glasgow, email@example.com, UK
- Kostas Kolomvatsos, University of Thessaly, firstname.lastname@example.org, Greece
- Stathes Hadjiefthymiades, University of Athens, email@example.com
- Arkady Zaslavsky, CSIRO, Arkady.Zaslavsky@csiro.au, Australia
- Alexandros Kalousis, University of Applied Sciences, firstname.lastname@example.org, Switzerland
- Konstantinos Oikonomou, Ionian University, email@example.com, Greece
You are kindly requested to address the following issues immediately,
as IEEE may be slow in responding to help requests.
Important Notice for Authors of Accepted Papers:
All Accepted Papers are required to pass IEEE PDF eXpress PDF Check by JULY 20, 2016 (Firm Deadline) in order to be included in the IEEE Xplore-compliant CD
Step by Step Instructions:
Access the IEEE PDF eXpress site at http://www.pdf-express.org (open June 23, 2016)
- First-time users: Click “New Users-Click Here”, enter 39952X for the Conference ID, your e-mail address and choose a new password. Continue to enter information as prompted. Check that the contact information is still valid and click Submit.
- Previous users, but using it the first time for a new conference: Enter 39952X for the Conference ID, your e-mail address and the password you used for your old account. When you click “Login”, you will receive a warning saying you need to set up an account. Simply click “Continue”. Enter your previously used e-mail address and password combination to enable your old account access the IISA2016 conference.
- Returning users: Enter 39952X for the Conference ID, e-mail address and password.
- For each conference paper, click “Create New Title”
- Enter identifying text for the paper (title is recommended but not required)
- Click “Submit PDF for Checking” or “Submit Source Files for Conversion”
- Indicate Platform, source file type (if applicable), click Browse and navigate to file, and click “Upload File”. You will receive online and e-mail confirmation of successful upload
- You will receive an e-mail with your Checked PDF or IEEE PDF eXpress-converted PDF attached. If you submitted a PDF for Checking, the email will show you if your file passed or failed.
AFTER you have successfully created the IEEE Xplore-compatible PDF file(s) you must re-upload the Final Camera-Ready file using Easychair Paper Submission System by JULY 20, 2016 (Firm Deadline)!
If the PDF submitted failes the PDF check:
Option 1: Submit your source file for conversion by clicking Try again, then Submit Source Files for Conversion.
Option 2: Read the PDF Check report, then click “The PDF Check Report” in the sidebar to get information on possible solutions.
Option 3: “Request Technical Help” through your account or via email at: firstname.lastname@example.org (include Conference id: 39952X).
If you are not satisfied with the IEEE PDF eXpress-converted PDF:
Option 1: Resubmit your source file with corrections Try again, then Submit Source Files for Conversion.
Option 2: Submit a PDF by clicking Try again, then Submit PDF for Checking
Option 3: “Request a Manual Conversion” through you account
General Conference Important Dates
Workshop | Special Session | Tutorial Proposals: February 29, 2016
April 11, 2016 May 20, 2016
Author Notification: May 27, 2016
Camera-Ready: June 06, 2016
Paper Submission Process
Note! During paper submission, when you reach the SELECT A TRACK page, if you are submitting a paper to a Workshop or Special Session, you are kindly requested to select the appropriate WS (Workshop Session) or SS (Special Session) from those listed on the paper submission form. Otherwise, please select GS (General Sessions)!
- Christos Anagnostopoulos, email@example.com
- Kostas Kolomvatsos, firstname.lastname@example.org