Marko Thaler: Difference between revisions
From Santa Fe Institute Events Wiki
mNo edit summary |
mNo edit summary |
||
Line 18: | Line 18: | ||
• invariant representation of data | • invariant representation of data | ||
• macro scale operation of biological neural networks | • macro-scale operation of biological neural networks | ||
Line 28: | Line 28: | ||
'''''4. Do you have any ideas for what sort of project you would like to attack this summer?''''' | '''''4. Do you have any ideas for what sort of project you would like to attack this summer?''''' | ||
• development of learning and prediction algorithms based on Hierarchical Temporal Memory framework | • development of learning and prediction algorithms based on [http://www.numenta.com/Numenta_HTM_Concepts.pdf Hierarchical Temporal Memory] framework | ||
• make a first fully autonomous flight of my UAV (Unmanned Aerial Vehicle) | • make a first fully autonomous flight of my UAV (Unmanned Aerial Vehicle) | ||
Line 48: | Line 48: | ||
In a kingdom far, far away lives a man by the name of Marko. He is a mechanical engineer and has always been interested in operation of human brain and the use of brain theory principles in human made intelligence systems. So in his graduate studies he set out on a journey to build a brain for an energy distribution system. The brain’s primary role would be to predict short term energy consumption of its clients from past experience and estimate the risk of prediction error. Based on this the system would determine its economically optimal operation. That would assure the system operating costs would be as low as possible while all the clients energy needs would be satisfied. After a successful birth of the first prediction brain an entrepreneurial bug bit him so he set out to use the prediction brains commercially in several district energy distribution systems. | In a kingdom far, far away lives a man by the name of Marko. He is a mechanical engineer and has always been interested in operation of human brain and the use of brain theory principles in human made intelligence systems. So in his graduate studies he set out on a journey to build a brain for an energy distribution system. The brain’s primary role would be to predict short term energy consumption of its clients from past experience and estimate the risk of prediction error. Based on this the system would determine its economically optimal operation. That would assure the system operating costs would be as low as possible while all the clients energy needs would be satisfied. After a successful birth of the first prediction brain an entrepreneurial bug bit him so he set out to use the prediction brains commercially in several district energy distribution systems. | ||
The developed brains were using simulated neural networks and genetic algorithms to solve their tasks. But Marko still was not satisfied with the intelligence capabilities his brains were having. So now he is starting to examine the use of more sophisticated brain theory principles to autonomously solve given problems. He is especially interested in using Jeff Hawkins’ memory-prediction framework of the brain (Hierarchical Temporal Memory) for intelligent extraction of information from past experience. To effectively use this framework he needs to have more knowledge in the fields of neuroscience, dynamics of complex systems, etc. Armed with this knowledge maybe some day he will be able to answer his favorite problem of how to develop an intelligent system that autonomously extracts and understands the information available on the internet. | The developed brains were using simulated neural networks and genetic algorithms to solve their tasks. But Marko still was not satisfied with the intelligence capabilities his brains were having. So now he is starting to examine the use of more sophisticated brain theory principles to autonomously solve given problems. He is especially interested in using Jeff Hawkins’ memory-prediction framework of the brain ([http://www.numenta.com/Numenta_HTM_Concepts.pdf Hierarchical Temporal Memory]) for intelligent extraction of information from past experience. To effectively use this framework he needs to have more knowledge in the fields of neuroscience, dynamics of complex systems, etc. Armed with this knowledge maybe some day he will be able to answer his favorite problem of how to develop an intelligent system that autonomously extracts and understands the information available on the internet. | ||
Whenever Marko is tired of playing with his artificial intelligence buddies he loves to interact and communicate with other people. In his free time he also practices some weird sport of golf, runs a lot and tries to devote as much time as possible to his princess. | Whenever Marko is tired of playing with his artificial intelligence buddies he loves to interact and communicate with other people. In his free time he also practices some weird sport of golf, runs a lot and tries to devote as much time as possible to his princess. |
Latest revision as of 12:59, 29 May 2006
DISCUSSION QUESTIONS
1. What topics do you have some expertise in and would you be willing to help others learn them?
• empirical modeling of complex systems (nonparametric regression, neural networks, genetic algorithms)
• risk estimation
• self organization of data (maximum entropy principle)
2. What do you want to learn?
• network dynamics and scaling
• invariant representation of data
• macro-scale operation of biological neural networks
3. Do you have any projects that would benefit from interdisciplinary approach?
• pattern extraction and correlation search in multi-component empirical data for developing better empirical models of complex energy distribution system
4. Do you have any ideas for what sort of project you would like to attack this summer?
• development of learning and prediction algorithms based on Hierarchical Temporal Memory framework
• make a first fully autonomous flight of my UAV (Unmanned Aerial Vehicle)
5. What's your favorite "big problem"?
• creation of methodology for autonomous extraction and understanding of information available on the internet
6. If you were given the opportunity to see where we were in one hundred years with respect to progress on one problem/subject, what would it be?
• human colonization of other planets
THE STORY
In a kingdom far, far away lives a man by the name of Marko. He is a mechanical engineer and has always been interested in operation of human brain and the use of brain theory principles in human made intelligence systems. So in his graduate studies he set out on a journey to build a brain for an energy distribution system. The brain’s primary role would be to predict short term energy consumption of its clients from past experience and estimate the risk of prediction error. Based on this the system would determine its economically optimal operation. That would assure the system operating costs would be as low as possible while all the clients energy needs would be satisfied. After a successful birth of the first prediction brain an entrepreneurial bug bit him so he set out to use the prediction brains commercially in several district energy distribution systems.
The developed brains were using simulated neural networks and genetic algorithms to solve their tasks. But Marko still was not satisfied with the intelligence capabilities his brains were having. So now he is starting to examine the use of more sophisticated brain theory principles to autonomously solve given problems. He is especially interested in using Jeff Hawkins’ memory-prediction framework of the brain (Hierarchical Temporal Memory) for intelligent extraction of information from past experience. To effectively use this framework he needs to have more knowledge in the fields of neuroscience, dynamics of complex systems, etc. Armed with this knowledge maybe some day he will be able to answer his favorite problem of how to develop an intelligent system that autonomously extracts and understands the information available on the internet.
Whenever Marko is tired of playing with his artificial intelligence buddies he loves to interact and communicate with other people. In his free time he also practices some weird sport of golf, runs a lot and tries to devote as much time as possible to his princess.
Now he is setting out on a journey half across the world to a beautiful small town of Santa Fe to conquer new knowledge, form new friendships and have an unforgettable experience.
To be continued…