arXiv subjects
Publications and source records attributed to Enrico Carlini.
We study the geometric structure of the statistical models for two-by-two contingency tables. One or two odds ratios are fixed and the corresponding models are shown to be a portion of a ruled quadratic surface or a segment. Some pointers to the general case of two-way contingency tables are also given and an application to case-control studies is presented.
In this paper, we consider two basic questions about presenting a homogeneous polynomial f: how many variables are needed for presenting f? How can one find a presentation of f involving as few variables as possible? We give a complete answer to both questions, determining the minimal number of variables needed, NEssVar(f), and describing these variables through their linear span, EssVar(f). Our results give rise to effective algorithms which we implemented in the computer algebra system CoCoA.
We reconsider the classical problem of representing a finite number of forms of degree d in n+1 variables as sums of powers of linear forms. We define a geometric construct called a `grove', which, in a number of cases allows us to determine the dimension of the space of forms which can be so represented. We also present two new `exceptional' examples, where this dimension is less than what a naive parameter count would predict.