Wyniki wyszukiwania - BazTech

1

Conditional mean embedding and optimal feature selection via positive definite kernels

Jorgensen Palle E.T., Song Myung-Sin, Tian James

Opuscula Mathematica

|

2024

|

Vol. 44, no. 1

79--103

EN

Motivated by applications, we consider new operator-theoretic approaches to conditional mean embedding (CME). Our present results combine a spectral analysis-based optimization scheme with the use of kernels, stochastic processes, and constructive learning algorithms. For initially given non-linear data, we consider optimization-based feature selections. This entails the use of convex sets of kernels in a construction of optimal feature selection via regression algorithms from learning models. Thus, with initial inputs of training data (for a suitable learning algorithm), each choice of a kernel K in turn yields a variety of Hilbert spaces and realizations of features. A novel aspect of our work is the inclusion of a secondary optimization process over a specified convex set of positive definite kernels, resulting in the determination of “optimal” feature representations.

2

Performance Evaluation of MC-CDMA Systems with Single User Detection Technique using Kernel and Linear Adaptive Method

Fateh Rachid, Darif Anouar, Safi Said

Journal of Telecommunications and Information Technology

|

2021

|

nr 4

1--11

EN

Among all the techniques combining multi-carrier modulation and spread spectrum, the multi-carrier code division multiple access (MC-CDMA) system is by far the most widely studied. In this paper, we present the performance of the MC-CDMA system associated with key single-user detection techniques. We are interested in problems related to identification and equalization of mobile radio channels, using the kernel method in Hilbert space with a reproducing kernel, and a linear adaptive algorithm, for MC-CDMA systems. In this context, we tested the efficiency of these algorithms, considering practical frequency selective fading channels, called broadband radio access network (BRAN), standardized for MC-CDMA systems. As far as the equalization problem encountered after channel identification is concerned, we use the orthogonality restoration combination (ORC) and the minimum mean square error (MMSE) equalizer techniques to correct the distortion of the channel. Simulation results demonstrate that the kernel algorithm is efficient for practical channel.

3

Detecting Insider Malicious Activities in Cloud Collaboration Systems

Smrithy G. S., Cuzzocrea A., Balakrishnan R.

Fundamenta Informaticae

|

2018

|

Vol. 161, nr 3

299--316

EN

Cloud Collaboration Systems (CCS) offer efficient coordination among users to work on shared tasks in diverse distributed environments such as social networking, healthcare, wikis, and intelligent systems. Many cloud collaboration systems services are basically loosely coupled in nature. The flexibility of such CCS lead to various vulnerabilities in the system since the users are given broad access privileges. This may result in catastrophic activities from malicious insiders which in turn result in major misuse and abuse of information. While many sophisticated security mechanisms have been established to detect outsider threats in various systems, a very few works have been reported so far to detect anomalous insider activities in complex CCS. In this paper, we propose a Sliding Window based Anomaly Detection using Maximum Mean Discrepancy or SWAD-MMD model to detect anomalous insider activities via access network of users and objects. The main scope of this paper is to exploit information theoretic and statistical techniques to address the above security issues in order to provide information theoretically provable security (i.e., anomaly detection with vanishing probability of error) based on graph based Maximum Mean Discrepancy (MMD) that measures the distance between mean embedding of distributions into a Reproducing Kernel Hilbert Space (RKHS). The theoretical aspects show that the proposed approach is suitable for detecting anomalous insider activities in dynamic cloud collaborative systems. Finally we validate the proposed model using two publicly available datasets from Wikipedia and present a performance evaluation in terms of accuracy of the proposed model.

4

Functional models for Nevanlinna families

Behrndt J., Hassi S., Snoo H. de

Opuscula Mathematica

|

2008

|

Vol. 28, no. 3

233-245

EN

The class of Nevanlinna families consists of R-symmetric holomorphic multivalued functions on C \ R with maximal dissipative (maximal accumulative) values on C+ (C-, respectively) and is a generalization of the class of operator-valued Nevanlinna functions. In this note Nevanlinna families are realized as Weyl families of boundary relations induced by multiplication operators with the independent variable in reproducing kernel Hilbert spaces.