No label defined (Q4336488)

The author considers the inverse mixed data problem: For a set of distinct real numbers \(\{\lambda_i\}^n_1\) and \((n-1)\) real numbers \(a_1,a_2,\dots,a_{[n/2]}\), \(b_1,b_2,\dots,b_{(n-1)/2]}\), construct an \(n\times n\) Jacobi matrix \(J\) such that \(J\) has eigenvalues \(\{\lambda_i\}^n_1\) and \(a_i=e^T_i Je_i\), \(i=1,2,\dots,[n/2]\); \(b_i=e^T_i Je_{i+1}\), \(i=1,2,\dots,[(n-1)/2]\), where \([x]\) denotes the greatest integer less than or equal to \(x\) and \(e_i\) is the \(i\)th colunm of the \(n\times n\) identity matrix. Based on the Cayley-Hamilton theorem, a necessary and sufficient condition for the existence of a solution of the mixed data problem is derived. From the proof of the condition, a method for finding the solution is presented and an example is given. The method is simple in construction and effective in calculation.