Abstract
Standard-form transformation is a technique for transforming a discrete ill-posed problem in general form (with a seminorm as regularizing term) into a standard-form problem. We give a simple geometric explanation why the weighted pseudoinverse is the correct inverse to use in the standard-form transformation. Our presentation is based on oblique projections and oblique pseudoinverses - tools that are often overlooked in numerical analysis.
Original language | English |
---|
Publication status | Published - 2004 |
---|