typos

Author: jonhare

docs/handouts/intro-handouts.pdf

@@ -99,7 +99,7 @@
 \begin{tabular}{ll}
 Data & $\{\bm{x}_n, \bm{y}_n\}^N_{n=1} \qquad \{\bm{x}_n\}^N_{n=1}$
 \vspace{3mm} \\ \pause
-Function Approximator & $\bm{y} = f (\bm{x}, \bm{\theta})$ %  + \nu
+Function Approximator & $\bm{y} = f (\bm{x}; \bm{\theta})$ %  + \nu
 \vspace{3mm} \\ \pause
 Parameter Estimation & $E_0 = \sum^N_{n=1} \{\|\bm{y}_n - f (\bm{x}_n; \bm{\theta})\|\}^2$
 \vspace{3mm} \\ \pause
@@ -121,11 +121,11 @@
 
 Deep learning is primarily characterised by function compositions: \\ \vspace{10mm}
 \begin{itemize}
-	\item<2-> Feedforward networks: $\bm{y} = f (g(\bm{x}, \bm\theta_g), \bm{\theta_f})$
+	\item<2-> Feedforward networks: $\bm{y} = f (g(\bm{x}; \bm\theta_g); \bm{\theta_f})$
 	\begin{itemize}
-		\item Often with relatively simple functions (e.g. $f(\bm x, \bm{\theta}_f) = \sigma(\bm{x}^\top \bm{\theta}_f)$)
+		\item Often with relatively simple functions (e.g. $f(\bm x; \bm{\theta}_f) = \sigma(\bm{x}^\top \bm{\theta}_f)$)
 	\end{itemize} \vspace{3mm}
-	\item<3-> Recurrent networks: $\bm y_t = f(\bm y_{t-1}, \bm x_t, \bm\theta) = f(f(\bm y_{t-2}, \bm x_{t-1}, \bm\theta), \bm x_t, \bm\theta) = \dots$
+	\item<3-> Recurrent networks: $\bm y_t = f(\bm y_{t-1}, \bm x_t; \bm\theta) = f(f(\bm y_{t-2}, \bm x_{t-1}; \bm\theta), \bm x_t; \bm\theta) = \dots$
 \end{itemize}
 \vspace{10mm}
 
@@ -289,7 +289,7 @@
 				\item You'll be using PyTorch as the primary framework, with Torchbearer to help out.
 				\item You will need to utilise GPU-compute for the later labs (we provide Google Colab links so you can use NVidia K80s or newer in the cloud).
 			\end{itemize}
-			\item Labs are in-person (Zepler L3) with a team of PhD student demonstrators \& myself and/or Antonia.
+			\item Labs are in-person (Zepler L3) with a team of PhD student demonstrators \& both of us.
 			\item Please ask lots of questions and use this time to get help on the labs and coursework.
 			\item After each lab you will have to do a follow-up problem-sheet exercise that will be marked.
 		\end{itemize}