We give an algorithm to PAC-learn the coecients of a mul-tivariate polynomial from the signs of its values, over a sample of real points which are only known approximately. While there are several p-pers dealing with PAC-learning polynomials (e.g. [3,11]), they mainly only consider variables over nite elds or real variables with no round-o error. In particular, to the best of our knowledge, the only other work considering rounded-o real data is that of Dennis Cheung [6]. There, multivariate polynomials are learned under the assumption that the co-ecients are independent, eventually leading to a linear programming problem. In this paper we consider the other extreme: namely, we consider the case where the coecients of the polynomial are (polynomial) functions of a single parameter.
CITATION STYLE
Shirley Cheung, H. C. (2000). Learning from approximate data. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 1858, pp. 407–415). Springer Verlag. https://doi.org/10.1007/3-540-44968-x_40
Mendeley helps you to discover research relevant for your work.