Wyniki wyszukiwania - BazTech

1

Automated approach to classification of mine-like objects using multiple-aspect sonar images

Wang X., Liu X., Japkowicz N., Matwin S.

Journal of Artificial Intelligence and Soft Computing Research

|

2014

|

Vol. 4, No. 2

133--148

EN

In this paper, the detection of mines or other objects on the seabed from multiple side-scan sonar views is considered. Two frameworks are provided for this kind of classification. The first framework is based upon the Dempster–Shafer (DS) concept of fusion from a single-view kernel-based classifier and the second framework is based upon the concepts of multi-instance classifiers. Moreover, we consider the class imbalance problem which is always presents in sonar image recognition. Our experimental results show that both of the presented frameworks can be used in mine-like object classification and the presented methods for multi-instance class imbalanced problem are also effective in such classification.

2

How To Construct Support Vector Machines Without Breaching Privacy

Zhan J., Chang L., Matwin S.

Studia Informatica : systems and information technology

|

2006

|

Vol. 1(7)

233--244

EN

This paper addresses the problem of data sharing among multiple parties in the following scenario: without disclosing their private data to each other, multiple parties, each having a private data set, want to collaboratively construct support vector machines using a linear, polynomial or sigmoid kernel function. To tackle this problem, we develop a secure protocol for multiple parties to conduct the desired computation. In our solution, multiple parties use homomorphic encryption and digital envelope techniques to exchange the data while keeping it private. All the parties are treated symmetrically: they all participate in the encryption and in the computation involved in learning support vector machines.

3

A Multi-Party Scheme for Privacy : Preserving Clustering

Zhan J., Matwin S., Chang L.

Studia Informatica : systems and information technology

|

2006

|

Vol. 1(7)

217--232

EN

Preserving data privacy while conducting data clustering among multiple parties is a demanding problem. We address this challenging problem in the following scenario: without disclosing their private data to each other, multiple parties, each having a private data set, want to collaboratively conduct k-medoids clustering. To tackle this problem, we develop secure protocols for multiple parties to achieve this dual goal. The solution is distributed, i.e., there is no central, trusted party having access to all the data. Instead, we define a protocol using homomorphic encryption and digital envelope techniques to exchange the data while keeping it private.