diff options
Diffstat (limited to 'report')
| -rwxr-xr-x | report/paper.md | 17 |
1 files changed, 5 insertions, 12 deletions
diff --git a/report/paper.md b/report/paper.md index 134243b..1df9335 100755 --- a/report/paper.md +++ b/report/paper.md @@ -156,8 +156,8 @@ K=1, as visible in figure \ref{fig:k-diff}. \begin{figure} \begin{center} -\includegraphics[width=17em]{fig/kneighbors_diffk.pdf} -\caption{NN recognition accuracy varying K. Split: 80-20} +\includegraphics[width=19em]{fig/kneighbors_diffk.pdf} +\caption{NN Accuracy varying K. Split: 80-20} \label{fig:k-diff} \end{center} \end{figure} @@ -277,8 +277,6 @@ after each step. # Question 3, LDA Ensemble for Face Recognition, PCA-LDA -In this section we will perform PCA-LDA recognition with NN classification. - Varying the values of $M_{\textrm{pca}}$ and $M_{\textrm{lda}}$ we obtain the average recognition accuracies reported in figure \ref{fig:ldapca_acc}. Peak accuracy of 93% can be observed for $M_{\textrm{pca}}=115$, $M_{\textrm{lda}}=41$; howeverer accuracies above 90% can be observed for $130 > M_{\textrm{pca}} 90$ and $50 > M_{\textrm{lda}} > 30$ values between 30 and 50. @@ -289,7 +287,7 @@ are displayed in table \ref{tab:time}. \begin{figure} \begin{center} -\includegraphics[width=17em]{fig/ldapca3dacc.pdf} +\includegraphics[width=20em]{fig/ldapca3dacc.pdf} \caption{PCA-LDA Accuracy when varying hyper-parameters} \label{fig:ldapca_acc} \end{center} @@ -376,7 +374,7 @@ Bagging is performed by generating each dataset for the ensembles by randomly pi \begin{figure} \begin{center} -\includegraphics[width=22em]{fig/bagging.pdf} +\includegraphics[width=20em]{fig/bagging.pdf} \caption{Ensemble size effect on accuracy with bagging} \label{fig:bagging-e} \end{center} @@ -392,7 +390,7 @@ use the 90 eigenvectors with biggest variance and picking 70 of the rest non-zer \begin{figure} \begin{center} -\includegraphics[width=23em]{fig/random-ensemble.pdf} +\includegraphics[width=21em]{fig/random-ensemble.pdf} \caption{Ensemble size - feature randomisation ($m_c=90$,$m_r=70$)} \label{fig:random-e} \end{center} @@ -509,8 +507,6 @@ We know that $S\boldsymbol{u\textsubscript{i}} = \lambda \textsubscript{i}\bolds From here it follows that AA\textsuperscript{T} and A\textsuperscript{T}A have the same eigenvalues and their eigenvectors follow the relationship $\boldsymbol{u\textsubscript{i}} = A\boldsymbol{v\textsubscript{i}}$ -### Table of execution times of different methods - \begin{table}[ht] \centering \begin{tabular}[t]{llll} @@ -522,12 +518,9 @@ PCA-ALT & 1.0 & 1.3 & 1.1 \\ LDA & 5.0 & 5.8 & 5.2 \\ LDA-PCA & 0.11 & 0.19 & 0.13 \\ \hline \end{tabular} -\caption{Comparison of execution times between different methods} \label{tab:time} \end{table} -### Memory Usage for NN and alternative method - \begin{figure} \begin{center} \includegraphics{fig/memnn.pdf} |