General Information • Project Description • Project Log • Links

General Information

Project Description

Traditional machine learning approaches make the assumption that instances are independent and identically distributed (IID). We term models constructed under the IID assumption first-order because in general they only leverage relationships between attributes within instances (e.g., co-occurrence relationships). Thus classification of a single instance (of previously unseen data) is possible because no additional context is needed to infer class membership. Such a context-free approach, however, does not exploit valuable information about relationships between instances in the dataset.

In our research we are developing a novel framework for learning that, unlike approaches that assume instances are IID, leverages implicit co-occurrence relationships between attributes in different instances. We term these implicit co-occurrence relationships higher-order paths. Attributes ( e.g., words in documents in text collections) are richly connected by such higher-order paths, and the model builts by our higher order learners exploit this rich connectivity pattern. In our work to-date we have developed both supervised and unsupervised learning approaches including Higher Order Naive Bayes, Higher Order SVM, Higher Order Classification Based ARM and Distributed Higher Order ARM. We are also have a framework under development that leverages human-computer interaction entitled Distributed Interactive Higher Order Privacy Enhancing Knowledge Discovery (DI HOPE KD).

Final Presentation (ppt)

Project Log

Week 1: Became associated with the problem and began to look at several previous studies in depth, focusing mainly on one produced by Microsoft.


Week 2: Started work on more robust forms of secret questions and made a presentation on proposed project all the while doing more research on different possible forms of secret questions.



Week 3: Finalized and developed a survey with the three proposed forms of secret questions and began to write up my findings.



Week 4: Conducted a small survey on four persons with the proposed secret questions. Wrote up the results and added them to my findings.



Week 5: Worked on journal article and further analyzed data from survey. Came up with a logical way of presenting findings within article and developed further thoughts concerning future studies. Visited Telecordia.



Week 6: The article was given an extention and as a result more has been added to it along with more editing. Future studies section was added and sections were moved. Started to look at the previous Microsoft Study and began to come up with ways in which to compare our findings with theirs, all mainly ended in conduting another survey. Visited Bell Labs.



Week 7: Prepared and presented my work in final presentation to the rest of the program.



Week 8: Final week of program. Wrapped up project and spoke with mentor about future studies.

Links and Resources

Rutgers DIMACS
Jodange
Its no secret; Measuring the security and reliability of authentication via secret question (pdf)
A Cryptographic Provenance Verification Approach For Host-Based Malware Detection (pdf)