Linear Regression from Uncertain Data and its Applications to Housing Price Prediction

Abstract

Most of the figures we get in our daily life are imprecise, a person's height is an approximation. Uncertain data is usually represented by many values around it in a limited region with probability density function. This paper first defines the uncertain data, and then proposes a linear regression algorithm based on the uncertain data. In order to verify the rationality, correctness and effectiveness of the algorithm, many experiments were carried out.