This paper presents a Semi-Supervised Feature Selection Method based on a univariate relevance measure applied to a multiobjective approach of the problem. Along the process of decision of the optimal solution within Pareto-optimal set, atempting to maximize the relevance indexes of each feature, it is possible to determine a minimum set of relevant features and, at the same time, to determine the optimal model of the neural network.

Multi-Objective Semi-supervised Feature Selection and Model Selection based on Pearson's Correlation Coefficient

/pdf/multi-objective-semi-supervised-feature-selection-and-model-1ngqftm6du.pdf

Multi-objective semi-supervised feature selection and model selection based on Pearson's correlation coefficient

Leakage-resilient (LR) cryptography including LR authenticated key exchange has been extensively studied in recent years. However, there is few literature to model and construct LR authenticated group key exchange (AGKE) protocol, which is the most practical cryptographic primitive for the group communication applications. In this paper, we first introduce an \(\lambda \)-bounded after-the-fact LR (\(\lambda \)-BAFLR) CK security model to assess the security of AGKE protocols in the leakage environments. Then by appropriately combining Diffie-Hellman (DH) group key exchange protocol and public key encryption with the security against adaptively chosen plaintext after-the-fact leakage attacks, we propose the first LR AGKE protocol, and show a formal proof of the \(\lambda \)-BAFLR CK security in the standard model based on the game simulation techniques. Our result shows that the proposed AGKE protocol attains the property of LR without incurring additional communication and computation cost by comparing with the related non-LR schemes.

An Efficient Leakage-Resilient Authenticated Group Key Exchange Protocol

This paper presents an efficient solution to key exchange problem for large groups in the password- based scenario. Our scheme can combine existing group protocols to get a hybrid group key exchange protocol which is efficient in terms of both computation and communication. And the resulting protocol allows users to securely join and leave the group at any time—the so-called dynamic case. For dynamic group communications, we propose an efficient key-updating algorithms so that the group key is only known to the members of the newly formed pool. Furthermore, our proposal does not introduce new idealizing assumptions or high entropy secrets for authentication.

Efficient Solution to Password-based Key Exchange for Large Groups

During the past two decades, side-channel attacks have become a familiar method of attacking cryptographic systems, which allow an attacker to learn partial information about the internal secrets such as the secret key. A scheme that is secure in the traditional model will be vulnerable in the leakage environments, thus designing a strong, meaningful, and achievable security scheme to capture the practical leakage attacks is one of the primary goals of leakage-resilient cryptography. In this work, we first formalize a continuous after-the-fact (AF) security model for leakage-resilient (LR) group password-authenticated key exchange (GPAKE) protocol, where the leakages are continuous and are allowed even after the adversary is given the challenges. Then, by combining Diffie-Hellman group key exchange protocol and Dziembowski-Faust leakageresilient storage scheme appropriately, we propose the first LR GPAKE protocol and present a formal security proof in the standard model.

/pdf/continuous-after-the-fact-leakage-resilient-group-password-3k4phu5v9q.pdf

Continuous After-the-Fact Leakage-Resilient Group Password-Authenticated Key Exchange.

This paper presents an efficient solution to the group key exchange problem in the password-based scenario. Our scheme can combine existing group protocols to get a hybrid group key exchange protocol which is efficient in terms of both computation and communication when the number of group users is very large. Our solution allows users to securely join and leave the group at any time--the so-called dynamic case. Moreover, we prove its security in the random oracle model.

Yuefei Zhu

Papers

Efficient Hybrid Password-Based Authenticated Group Key Exchange