International Journal of Application or Innovation in Engineering & Management (IJAIEM)
Random projections are a powerful method of dimensionality reduction that are noted for their simplicity and strong error guarantees. In this research paper, the authors study a randomized multiplicative data perturbation technique for privacy preserving data mining. It is motivated by the work presented earlier that point out some security problems of additive perturbation and distance preserving perturbation. They provide a theoretical result related to projections and explores the possibility of using multiplicative random projection matrices for constructing a new representation of the data.