Neurocomputation Laboratory

L. M. Manevitz, Director of Laboratory

1. Goals and Background

The field of Neurocomputation involves the interaction of an analysis of way the human brain works and the theory and practice of computation.

As such, it is an interdisciplinary field, calling on and and affecting: computer science, brain science (physiology and psychology), mathematics, statistics, physics and philosophy.

It has two main research aims, both of which are being pursued at the Laboratory:

• On the one hand, it is concerned with the possibility of computation in computers by following the paradigm and analysis of computation that occurs in neurons and the brain. In recent years, this has resulted in breakthroughs in pattern recognition, learning theory, clustering, associative memory and fault tolerant computation. The most successful adaptive computing implementations today use this ”neural network” technology. (For example, credit card fraud is tracked in this manner).


• On the other hand, the precision resulting from the computational and mathematical viewpoint has resulted in insights helping to clarify one of the ultimate human research endeavors: understanding the way the human brain works.


• The Neurocomputation Laboratory is dedicated to pursuing these studies by supporting and encouraging research, education and technology transfer in these fields.

2. Recent Activities

2.1 Research Activities

Here are some of the research topics being pursued at the Laboratory.

• Personalized, Automated, and Adaptive Information Retrieval over the Internet
• Models of Neuronal Activity
• Automatic Adaptive and Predictive Meshing in the Finite Element Method
• Modeling Dual Task Performance in Humans
• Automatic Assigning Geometry to Meshes in the Finite Element Method
• Simplified Models of the Neocortex
• Relationship between Discrete and Continuous Neuronal Models
• Philosophical Analysis of PsychoPathology: Practical Applications

2.2 Education

• Under a grant of CRI, a new course, Mathematical Models of Neuronal Activity is being prepared and will be given in the Spring, 2004 semester.
• The interdisciplinary group, designed to develop ”modules” for interdisciplinary courses in related to the field of Brain Science has been funded by a CRI grant and will be meeting under the auspices of the Laboratory
• An interdisciplinary seminar in the subject of Computation and the Mind is sponsored by the Laboratory.
• The Laboratory encourages and will give minor support to other courses and research seminars, such as the course ”Neurocomputation” and ”Advanced Topics Seminar in Neurocomputation” in the Computer Science Department, and the Brain and Behavior Seminar (run by Gal Richter- Levin of the Psychology Department in conjunction with the Technion.)

2.3 Research Meetings and Research Guests

• International Meetings

The Laboratory has participated in the organization of under the auspices of CRI, major international conferences:

1. Mathematical Problems (and Solutions?) Arising from Neurophysiology and
2. What is Learning in the Neural System?
3. The 16th Meeting of the Israeli Society for Computational Mechanics: Computational Mechanics and Soft Computing


• Guests

The Laboratory hosts lectures and research guests, some associated with the above meetings; others coming for direct research visits. Included in the above is Nobel Prize Laureate Bert Sakmann. Planned guests (not all confirmed) for 2003-2004 include Pablo Padilla of Mexico, Philip Maini of Oxford U., Leonardo Franco of Oxford U., and Roy Raj of Cranfield U., David Khardon of Southhampton U.

2.4 Technology Transfer

The laboratory encourages the transfer of research to industry.

• A patent, under the auspices of the University Technology Center, ”Carmel” has been filed concerning ”Adaptive Meshing in the Finite Element Method”.
• Initial considerations regarding the patentability of ideas related to neural networks and music are underway.

2.5 Associated Research Members of the Laboratory

These are researchers, mostly located at other universities, who have a close tie with the Laboratory and make frequent visits.

1. Akram Bitar, IBM Research, Haifa
2. Dan Givoli, Dept. of Aerospace Engineering, Technion
3. Shimon Marom, Dept. of Physiology, Technion
4. Malik Yousef, Informatics Laboratory, U. Pennsylvania

2.6 Recent Advanced Degree Students and Post-Doctorates

• Irit Shreir, Post-Doc, Dept. of Philosophy, 2003
• Monash Khashkoosh, M.Sc. Dept. of Mathematics, 2003
• Akram Bitar, M.Sc., Dept of Mathematics and Computer Science, 2002
• Malik Yousef, Ph.D., Dept. of Mathematics and Computer Science, 2003

2.7 Recent Publications

1. Meltser, M., Shoham, M. and Manevitz, L. Approximating Functions by Neural Netw orks: A Constructive Solution in the Uniform Norm, Neural Networks, Vol. 9, 965-978.
2. Manevitz, L., Givoli, D. and Margi, M. Heuristic finite element node numbering: an expert system approach, Computing Systems in Engineering Vol. 4,p. 159-168.
3. Manevitz, L., Yousef, M. and Givoli, D. Finite Element Mesh Generation Using Self-Organizing Neural Networks, Special Issue on Machine Learning of MicroComputers in Civil Engineering, Vol. 12 No. 4, 233 - 250 .
4. Manevitz, L. Interweaving Kohonen Maps of Different Dimensions to Handle Measure Zero Constraints on Topological Mappings, Neural Processing Letters, Vol. 5 No. 2, 155-161.
5. Manevitz, L. and Givoli, D. Automating the Finite Element Method: A Test-Bed for Soft Computing, Annals of Soft Computing, 2003.
6. Manevitz, L., Bitar, A. and Givoli, D. Finite Element Mesh Adaptation via Time Series Prediction and Neural Networks, submitted, Neurocomputing.
7. Manevitz, L. and Marom, S. Modeling the Process of Rate Selection in Neuronal Activity, J. Theoretical Biology, 2002.
8. Manevitz, L. and Yousef, M., A Web Navigation System Based on a Neural Network User-Model Trained with Only Positive Web Documents, WIAS, to appear , 2003.