Sparse Coding
From Ufldl
(→Probabilistic Interpretation) |
(→Probabilistic Interpretation [Based on Olshausen and Field 1996]) |
||
Line 93: | Line 93: | ||
\end{align}</math> | \end{align}</math> | ||
- | Using a probabilistic approach, it can also be seen that the choices of the <math>L_1</math> penalty <math>\left|a_i\right|_1 </math> and the log penalty <math>\log(1+a_i^2)</math> for <math>S(.)</math> correspond to the use of the Laplacian | + | Using a probabilistic approach, it can also be seen that the choices of the <math>L_1</math> penalty <math>\left|a_i\right|_1 </math> and the log penalty <math>\log(1+a_i^2)</math> for <math>S(.)</math> correspond to the use of the Laplacian <math>P(a_i) \propto \exp\left(-\beta|a_i|\right)</math> and the Cauchy prior <math>P(a_i) \propto \frac{\beta}{1+a_i^2}</math> respectively. |
== Learning == | == Learning == |