University AI - Supervised Parametric Estimation Algorithms (Slides)

Differences between Non-Parametric and Parametric Estimation

• Parametric Estimation: we search for the parameters ($\underline{\Theta }$) that describe a PDF of an arbitrary form, so we can calculate $p(\underline{x}|{\omega }_i)$
• Non-Parametric Estimation: no assumption on the form of the PDF are made, we calculate $P({\omega }_i|\underline{x_0})$ given the sample data or we first calculate $p(\underline{x_0}|{\omega }_i)$, $P({\omega }_i)$ and $p(\underline{x_0})$ and and then use the Bayes theorem to find the posterior probability $P({\omega }_i|\underline{x_0})$.

---

Algorithms for Supervised Parametric Estimation