0
$\begingroup$

I completely understood the math behind PCA. I have a doubt here while calculating the function that will do the transformation. According to the book : Deep Learning by Ian Goodfellow, Yoshua Bengio and Aaron Courville :

enter image description here

But when we do Transformation, We go for T = XD instead of above. How both of these are similar. I am confused :(

$\endgroup$
0
$\begingroup$

In the book, $x\in \mathbb{R}^{d\times 1}$, but typically in code implementations/examples, $X$ is a data matrix with dimension $n\times d$, where each row is a data sample of dimension $1\times d$. Therefore, $$XD=\begin{bmatrix}x_1^T\\x_2^T\\\vdots \\ x_n^T\end{bmatrix}D=\begin{bmatrix}x_1^TD\\x_2^TD\\\vdots \\ x_n^TD\end{bmatrix}=\begin{bmatrix}f_1(x)^T\\f_2(x)^T\\\vdots\\f_n(x)^T\end{bmatrix}$$ In conclusion, it's the same thing transposed.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.