| Line 4: | Line 4: | ||
*Hello everybody, I am Prof. Mimi, your instructor. My goal in this course is to teach you some fundamental tools that you can use to attack the different decision theory problems that you will encounter, along with the limits of these tools. I also hope to give you some ideas on how to address the limitations of current decision theory techniques. I am personally very interested in decision theory. Part of my research focuses on designing methods to extract features to represent data in such a way that no information is lost in the process. I am also interested in developing methods to remove information from a representation that is not helpful in making a decision. This is particularly challenging when the underlying data is high-dimensional. Looking forward to getting to know you all! --[[User:Mboutin|Mboutin]] 15:33, 12 January 2010 (UTC) | *Hello everybody, I am Prof. Mimi, your instructor. My goal in this course is to teach you some fundamental tools that you can use to attack the different decision theory problems that you will encounter, along with the limits of these tools. I also hope to give you some ideas on how to address the limitations of current decision theory techniques. I am personally very interested in decision theory. Part of my research focuses on designing methods to extract features to represent data in such a way that no information is lost in the process. I am also interested in developing methods to remove information from a representation that is not helpful in making a decision. This is particularly challenging when the underlying data is high-dimensional. Looking forward to getting to know you all! --[[User:Mboutin|Mboutin]] 15:33, 12 January 2010 (UTC) | ||
| − | * | + | *Howdy, my name is Philip and I am a 2nd year PhD student in Computer Science working with Prof. Vernon Rego on generating and detecting hidden information in covert channels. Basically, I have two problems; the first is this: given a data set containing objects which may or may not contain hidden information (covert communication... think secret agents), find the objects which contain hidden information. The other problem is to generate objects within which we can hide information and do this in such a way that these new information-hiding objects can evade detection by the aforementioned detection system. I hope to learn from this course several methods for attacking my detection problem. I am looking forward to learning more about feature extraction and information representation methods. [[User:Pritchey|Pritchey]] 16:09, 12 January 2010 (UTC) |
Revision as of 12:09, 12 January 2010
Discussion Topic 1: Introduction and Expections
- Hello everybody, I am Prof. Mimi, your instructor. My goal in this course is to teach you some fundamental tools that you can use to attack the different decision theory problems that you will encounter, along with the limits of these tools. I also hope to give you some ideas on how to address the limitations of current decision theory techniques. I am personally very interested in decision theory. Part of my research focuses on designing methods to extract features to represent data in such a way that no information is lost in the process. I am also interested in developing methods to remove information from a representation that is not helpful in making a decision. This is particularly challenging when the underlying data is high-dimensional. Looking forward to getting to know you all! --Mboutin 15:33, 12 January 2010 (UTC)
- Howdy, my name is Philip and I am a 2nd year PhD student in Computer Science working with Prof. Vernon Rego on generating and detecting hidden information in covert channels. Basically, I have two problems; the first is this: given a data set containing objects which may or may not contain hidden information (covert communication... think secret agents), find the objects which contain hidden information. The other problem is to generate objects within which we can hide information and do this in such a way that these new information-hiding objects can evade detection by the aforementioned detection system. I hope to learn from this course several methods for attacking my detection problem. I am looking forward to learning more about feature extraction and information representation methods. Pritchey 16:09, 12 January 2010 (UTC)
