From: Mart Lubbers Date: Mon, 12 Sep 2016 18:30:19 +0000 (+0200) Subject: update X-Git-Url: https://git.martlubbers.net/?a=commitdiff_plain;h=8d38f5cb57c87fefced4d6c2c0dbc1cf491231ce;p=itlast1617.git update --- diff --git a/week3/eml.tex b/week3/eml.tex index 5e13726..781f2fe 100644 --- a/week3/eml.tex +++ b/week3/eml.tex @@ -2,7 +2,6 @@ \usepackage{amssymb} \usepackage{booktabs} -\usepackage{float} \usepackage{geometry} \title{Exercise: Machine Learning} @@ -12,7 +11,19 @@ \begin{document} \maketitle \subsection*{Chapter 5: Machine Learning} -\begin{table}[H] +Table~\ref{t2} shows that there is some difference in classification when +choosing different parameter sets. The results show that adding the focus word +is very important knowledge since the percentage increases quite a bit. +However, knowing only the focus word gives exceptionally low performance. + +Knowing the next words or the previous words gives some improvement but not a +whole lot. + +Table~\ref{t1} shows that the method does not make a big difference. + +Using ten-fold cross validation decreases the percentage. + +\begin{table} \centering \begin{tabular}{lll} \toprule @@ -24,38 +35,37 @@ J48 (10FCF) & $96.4352\%$ & $36.5122\%$\\ \bottomrule \end{tabular} - \caption{Results for \texttt{P1D} and \texttt{FD}} + \caption{Results for \texttt{P1D} and \texttt{FD}\label{t1}} \end{table} -\begin{table}[H] +\begin{table} \centering \begin{tabular}{lllllll} \toprule \texttt{P2D} & \texttt{P1D} & \texttt{N2D} & \texttt{N1D} & \texttt{FW} & Correctly classified\\ \midrule - \checkmark{} & \checkmark{} & \checkmark{} & \checkmark{} & \checkmark{} & $98.3225\%$\\ - \checkmark{} & \checkmark{} & \checkmark{} & \checkmark{} & & $97.3905\%$\\ - \checkmark{} & \checkmark{} & & \checkmark{} & \checkmark{} & $98.5555\%$\\ - \checkmark{} & \checkmark{} & & \checkmark{} & & $97.507\%$\\ - \checkmark{} & \checkmark{} & & & \checkmark{} & $98.0429\%$\\ - \checkmark{} & \checkmark{} & & & & $95.5732\%$\\ - - & \checkmark{} & \checkmark{} & \checkmark{} & \checkmark{} & $\%$\\ - & \checkmark{} & \checkmark{} & \checkmark{} & & $\%$\\ - & \checkmark{} & & \checkmark{} & \checkmark{} & $\%$\\ - & \checkmark{} & & \checkmark{} & & $\%$\\ - & \checkmark{} & & & \checkmark{} & $98.2992\%$\\ - & \checkmark{} & & & & $96.6449\%$\\ - - & & \checkmark{} & \checkmark{} & \checkmark{} & $\%$\\ - & & \checkmark{} & \checkmark{} & & $\%$\\ - & & & \checkmark{} & \checkmark{} & $\%$\\ - & & & \checkmark{} & & $\%$\\ - & & & & \checkmark{} & $88.4436\%$\\ - + \checkmark{} & \checkmark{} & \checkmark{} & \checkmark{} & \checkmark{} & $98.3225\%$\\ + \checkmark{} & \checkmark{} & \checkmark{} & \checkmark{} & & $97.3905\%$\\ + \checkmark{} & \checkmark{} & & \checkmark{} & \checkmark{} & $98.5555\%$\\ + \checkmark{} & \checkmark{} & & \checkmark{} & & $97.507\%$\\ + \checkmark{} & \checkmark{} & & & \checkmark{} & $98.0429\%$\\ + \checkmark{} & \checkmark{} & & & & $95.5732\%$\\ + + & \checkmark{} & \checkmark{} & \checkmark{} & \checkmark{} & $98.6486\%$\\ + & \checkmark{} & \checkmark{} & \checkmark{} & & $97.5769\%$\\ + & \checkmark{} & & \checkmark{} & \checkmark{} & $98.5555\%$\\ + & \checkmark{} & & \checkmark{} & & $97.973\%$\\ + & \checkmark{} & & & \checkmark{} & $98.2992\%$\\ + & \checkmark{} & & & & $96.6449\%$\\ + + & & \checkmark{} & \checkmark{} & \checkmark{} & $91.8919\%$\\ + & & \checkmark{} & \checkmark{} & & $85.5079\%$\\ + & & & \checkmark{} & \checkmark{} & $92.579\%$\\ + & & & \checkmark{} & & $85.2516\%$\\ + & & & & \checkmark{} & $88.4436\%$\\ \bottomrule \end{tabular} - \caption{NaiveBayes} + \caption{NaiveBayes on all sensible combinations\label{t2}} \end{table} \subsection*{Chapter 6: Exercises} @@ -65,10 +75,11 @@ the preceding $n$ states. However, it is claimed that the algorithm takes into account the whole sequence. Explain in your own words (at most $100$) how the probability is influenced by the rest of the - sequence, i.e.\ both the positions more than n back and the following + sequence, i.e.\ both the positions more than $n$ back and the following positions.} \item\emph{Explain in your own words (at most 50) how the EM algorithm works. I don't mean the mathematics, but the underlying concept.} + \end{itemize} \end{document}