Joint meeting with the "Distinguished Lecturer" - Prof. James Bezdek, USA

Share

https://softuni.bg/trainings/1411/ieee-summer-school-on-systems-man-and-cybernetics

https://softuni.bg/trainings/1405/computational-intelligence


The IEEE Fellow Prof. James Bezdek will be our guest within the “Distinguished Lecturer” program of IEEE Computational Intelligence Society in the beginning of this September. He will deliver a cycle of two talks entitled:

1. How big is too big? (Mostly) c-Means Clustering in Big Data.  (7th September)

2. Anomaly Detection in Wireless Sensor Networks: Visual Assessment and Clustering for Environmental Monitoring.  (3rd September)

 

He visits Bulgaria at the invitation of IEEE Young Professionals Affinity Group of Bulgaria and IEEE Computational Intelligence Chapter of Bulgaria with the support of IEEE Computational Intelligence Society.



  Date and Time

  Location

  Contact

  Registration


  • Start time: 03 September 2016 11:00 AM
  • End time: 07 September 2016 04:00 PM
  • All times are Europe/Sofia
  • Google_Cal_icon Add to Google Calendar
  • 15, Tintiava st.
  • Sofia, Sofiya-Grad
  • Bulgaria
  • Building: Software University
  • Click here for Map

Staticmap?size=250x200&sensor=false&zoom=14&markers=42.6666644%2c23
  • yancho.todorov@ieee.org

  • Co-sponsored by Software University


  Speakers

Prof. James Bezdek

Prof. James Bezdek

Topic:

How big is too big? (Mostly) c-Means Clustering in Big Data

What is big data? Objectives of clustering in big data are acceleration for loadable data and feasibility for non-loadable data. History of c-means and least squares estimation.  Acceleration methods for fuzzy c-means (FCM). Approximate clustering with FCM and Gaussian mixture models (EM) based on literal clustering of a sample followed by non-iterative extension. Incremental methods (spFCM and olFCM) that process data chunks sequentially. Extension of VAT to scalable VAT (sVAT) for arbitrarily large square data. sVAT marries single linkage (SL), resulting in two offspring: scalable SL and clusiVAT. Time and accuracy comparisons of clusiVAT to crisp versions of the three FCM models and CURE. Experiments on 48 synthetic data sets of Gaussian clusters, and three real world data sets (Iris, Forest and the KDD-99 cup). For example, clusiVAT recovers 97.06% of the crisp labels from the KDD-99 cup data,
a set of 4, 292,637 vectors, each having 41 attributes, in 76 seconds.

Email:

Address:Pensacola, Florida, United States

Prof. James Bezdek

Topic:

Anomaly Detection in Wireless Sensor Networks: Visual Assessment and Clustering for Environmental Monitoring.

Anomalies in Wireless Sensor Networks (WSNs). (i) Isolated and epoch anomalies internal to a node; aberrant behavior of an entire node; and anomalous subtrees. (ii) Models that use data capture by level sets of ellipsoids (iii) Models that use visual assessment of elliptical summaries (iv) Measures of (dis)similarity on sets of ellipsoids (v) Visual evidence for cluster tendency in sets of ellipsoids (vi) Numerical examples using single linkage clustering on real WSN data from the IBRL network, the Great Barrier Reef Ocean Observation System, and the Grand St. Bernard pass.

Biography:

Prof. James C. Bezdek is recognized as one of the most important researchers in the world in the field of fuzzy systems for pattern recognition. He is past president of NAFIPS (North American Fuzzy Information Processing Society), IFSA (International Fuzzy Systems Association) and the IEEE CIS (Computational Intelligence Society). He is founding editor of the International Journal on Approximate Reasoning and the IEEE Transactions on Fuzzy Systems. He is Fellow member of the IEEE and IFSA, and recipient of the IEEE 3rd Millennium, IEEE CIS Fuzzy Systems Pioneer, and IEEE CIS Rosenblatt medals.

He is author of the fuzzy c-means (FCM) algorithm, considered one of the most important discoveries in fuzzy pattern recognition and related areas and the clustering algorithm of choice for most practitioners in fuzzy exploratory data analysis. The original model has inspired many applications in related areas of pattern recognition and image processing.

Areas of research benefiting from Dr. Bezdek's work include diagnostic medicine, economics, chemistry, image processing, meteorology, web mining, geology, target recognition, regression analysis, document retrieval, structural failure and irrigation models. One of the most notable applications has been in medical image analysis, where FCM segmentation of magnetic resonance images is used in conjunction with rule-based analysis for both diagnosis and pre-operative planning for brain tumor patients. Dr. Bezdek also has made pioneering contributions in deriving the theories for clustering of relational (Euclidean and non-Euclidean) data.
Dr. Bezdek is Honorary Senior Fellow Professor at University of Melbourne, and the Nystul Professor and Eminent Scholar at the University of West Florida in Pensacola.
Jim's interests: woodworking, optimization, motorcycles, pattern recognition, cigars, clustering in very large data, fishing, poker, co-clustering, blues music, and visual clustering in relational data.

Email:

Address:Pensacola, Florida, United States

Prof. James Bezdek

Topic:

How big is too big? (Mostly) c-Means Clustering in Big Data

Biography:

Email:

Address:Pensacola, Florida, United States

Prof. James Bezdek

Topic:

Anomaly Detection in Wireless Sensor Networks: Visual Assessment and Clustering for Environmental Monitoring.

Biography:

Email:

Address:Pensacola, Florida, United States





Agenda

The talks are sheduled for 3rd and 7th of September 2016! The exact hours will be propagated in a later point of time! The talk on 7th September will be merged with the IEEE Summer School on SMC!

for more information: http://yp.ieee.bg