This paper proposes an identity-based threshold key escrow scheme. The scheme is secure against identity-based threshold chosen-plaintext attack. It tolerates the passive adversary to access data of corrupted key escrow agency servers and the active adversary that can modify corrupted servers’ keys. The formal proof of security is presented in the random oracle model, assuming the Bilinear Diff...

This paper proposes a threshold key escrow scheme from pairing. It tolerates the passive adversary to access any internal data of corrupted key escrow agents and the active adversary that can make corrupted servers to deviate from the protocol. The scheme is secure against threshold adaptive chosen-ciphertext attack. The formal proof of security is presented in the random oracle model, assuming...

The crisscross error model in data arrays is considered, where the corrupted symbols are confined to a prescribed number of rows or columns (or both). Under the additional assumption that the corrupted entries are uniformly distributed over the channel alphabet, and by allowing a small decoding error probability, a coding scheme is presented where the redundancy can get close to one half the re...

We study the problem of estimating the highdimensional Gaussian graphical model where the data are arbitrarily corrupted. We propose a robust estimator for the sparse precision matrix in the highdimensional regime. At the core of our method is a robust covariance matrix estimator, which is based on truncated inner product. We establish the statistical guarantee of our estimator on both estimati...

loud storage mainly offers an on-demand data outsourcing servicing model which is gaining popularity through its elasticity and low maintenance cost. In such cases Security concerns arises when data is outsourced to a third-party cloud storage provider. The data that is stored in the cloud might get corrupted due to some malicious programs or due to an adversary. Considering the existing method...

Many supervised learning problems involve learning from samples whose labels are corrupted in some way. For example, each label may be flipped with some constant probability (learning with label noise), or one may have a pool of unlabelled samples in lieu of negative samples (learning from positive and unlabelled data). This paper uses class-probability estimation to study these and other corru...