📄 page_165.html
字号:
<HTML> <HEAD> <!--SCRIPT LANGUAGE="JavaScript" SRC="http://a1835.g.akamai.net/f/1835/276/3h/www.netlibrary.com/include/js/dictionary_library.js"></SCRIPT> <SCRIPT LANGUAGE="JavaScript"> if (!opener){document.onkeyup=parent.turnBookPage;} </SCRIPT!--> <META HTTP-EQUIV="Cache-Control" CONTENT="no-cache"> <META HTTP-EQUIV="Pragma" CONTENT="no-cache"> <META HTTP-EQUIV="Expires" CONTENT="-1"><META http-equiv="Content-Type" content="text/html; charset=windows-1252"><SCRIPT>var PrevPage="Page_164";var NextPage="Page_166";var CurPage="Page_165";var PageOrder="174";</SCRIPT> <TITLE>Document</TITLE> </HEAD> <BODY BGCOLOR="#FFFFFF"><CENTER><TABLE BORDER=0 WIDTH=100% CELLPADDING=0><TR><TD ALIGN=CENTER> <TABLE BORDER=0 CELLPADDING=2 CELLSPACING=0 WIDTH=100%> <TR> <TD ALIGN=LEFT><A HREF='Page_164.html'>Previous</A></TD> <TD ALIGN=RIGHT><A HREF='Page_166.html'>Next</A></TD> </TR> </TABLE></TD></TR><TR><TD ALIGN=LEFT><P><A NAME='JUMPDEST_Page_165'/><A NAME='{5C7}'/><TABLE BORDER=0 CELLSPACING=0 CELLPADDING=0 WIDTH='100%'><TR><TD ALIGN=RIGHT><FONT FACE='Times New Roman, Times, Serif' SIZE=2 COLOR=#FF0000>Page 165</FONT></TD></TR></TABLE><A NAME='{5C8}'/><TABLE BORDER=0 CELLSPACING=0 CELLPADDING=0><TR> <TD ROWSPAN=5></TD> <TD COLSPAN=3 HEIGHT=12></TD> <TD ROWSPAN=5></TD></TR><TR> <TD COLSPAN=3></TD></TR><TR><TD></TD> <TD><FONT FACE='Times New Roman, Times, Serif' SIZE=3>Once a decision tree is created there are two ways it can be validated. The simplest and fastest way is by going to the <I>Test</I> menu and selecting the <I>Resubstitution </I> option, which provides you both a Misclassification Rate and an Accuracy figure. The other method of evaluating a decision tree is by splitting your data set by whatever ratio you want, and then testing it:</FONT></TD><TD></TD></TR><TR> <TD COLSPAN=3></TD></TR><TR> <TD COLSPAN=3 HEIGHT=1></TD></TR></TABLE><A NAME='{5C9}'/><TABLE BORDER=0 CELLSPACING=0 CELLPADDING=0><TR> <TD ROWSPAN=5></TD> <TD COLSPAN=3 HEIGHT=12></TD> <TD ROWSPAN=5></TD></TR><TR> <TD COLSPAN=3></TD></TR><TR><TD></TD> <TD><FONT FACE='Courier New, Courier, Mono New, Courier, Mono' SIZE=2>KnowledgeSEEKER test results of partitioned data set. <BR>Error Rate Profile :: Test Partition <BR> Learning Testing Dropped <BR> Category Freq % Freq % Diff. Freq % <BR>—— — — — — — | — <BR> NoSale.htm 776 70.3 254 69.0 -1.3 |0 0.0 <BR> Sale.htm 328 29.7 114 31.0 1.3 |0 0.0 <BR>—— — — — — —— |— <BR> 1104 368 0 <BR>Misclassification Rate: 0.290761 <BR>Accuracy: 70.9239 </FONT></TD><TD></TD></TR><TR> <TD COLSPAN=3></TD></TR><TR> <TD COLSPAN=3 HEIGHT=1></TD></TR></TABLE><A NAME='{5CA}'/><TABLE BORDER=0 CELLSPACING=0 CELLPADDING=0><TR> <TD ROWSPAN=5></TD> <TD COLSPAN=3 HEIGHT=12></TD> <TD ROWSPAN=5></TD></TR><TR> <TD COLSPAN=3></TD></TR><TR><TD></TD> <TD><FONT FACE='Times New Roman, Times, Serif' SIZE=3>Decision tree settings can be adjusted via the Tree Configuration menu. For example, the <I>Filter Threshold</I> (percent error) can be set to the following settings:</FONT></TD><TD></TD></TR><TR> <TD COLSPAN=3></TD></TR><TR> <TD COLSPAN=3 HEIGHT=1></TD></TR></TABLE><A NAME='{5CB}'/><TABLE CELLPADDING=0 CELLSPACING=0 BORDER=0 WIDTH='100%'><TR><TD HEIGHT=12></TD></TR><TR><TD><TABLE CELLSPACING=0 WIDTH=613 CELLPADDING=5><TR><TD WIDTH=181 VALIGN=TOP><FONT FACE='Times New Roman, Times, Serif' SIZE=2>Decision</FONT></TD><TD WIDTH=432 VALIGN=TOP><FONT FACE='Times New Roman, Times, Serif' SIZE=2>1%</FONT></TD></TR><TR><TD WIDTH=181 VALIGN=TOP><FONT FACE='Times New Roman, Times, Serif' SIZE=2>Prediction</FONT></TD><TD WIDTH=432 VALIGN=TOP><FONT FACE='Times New Roman, Times, Serif' SIZE=2>5%</FONT></TD></TR><TR><TD WIDTH=181 VALIGN=TOP><FONT FACE='Times New Roman, Times, Serif' SIZE=2>Exploration</FONT></TD><TD WIDTH=432 VALIGN=TOP><FONT FACE='Times New Roman, Times, Serif' SIZE=2>20%</FONT></TD></TR></TABLE></TD></TR></TABLE><BR><A NAME='{5CC}'/><TABLE BORDER=0 CELLSPACING=0 CELLPADDING=0><TR> <TD ROWSPAN=5></TD> <TD COLSPAN=3 HEIGHT=12></TD> <TD ROWSPAN=5></TD></TR><TR> <TD COLSPAN=3></TD></TR><TR><TD></TD> <TD><FONT FACE='Times New Roman, Times, Serif' SIZE=3>The <I>Decision</I> support setting of 1 percent means that erroneous effects will be produced only 1 time in 100. It is a setting requiring an exceptionally high level of confidence in tree results. The <I>Prediction</I> is the default setting confirming that the displayed relationships are valid with a 95 percent certainty rating. This is the recommended setting for most data mining tasks by ANGOSS. The <I>Exploration</I> setting is the most relaxed and is designed to let you explore relationships in your data. It allows you to be wrong in about 20 times out of 100 when using this filter level. Lastly, KnowledgeSEEKER gives you the option to specify your own filter threshold for any value between 0 and 1. Other tree configuration options include <I>Split Search Algorithm:</I></FONT><FONT FACE='Times New Roman, Times, Serif' SIZE=3 COLOR=#FFFF00><!-- break --></FONT></TD><TD></TD></TR><TR> <TD COLSPAN=3></TD></TR><TR> <TD COLSPAN=3 HEIGHT=1></TD></TR></TABLE><A NAME='{5CD}'/><TABLE CELLPADDING=0 CELLSPACING=0 BORDER=0 WIDTH='100%'><TR><TD HEIGHT=12></TD></TR><TR><TD><TABLE CELLSPACING=0 WIDTH=583 CELLPADDING=5><TR><TD WIDTH=151 VALIGN=TOP><FONT FACE='Times New Roman, Times, Serif' SIZE=2><I>Cluster</I></FONT></TD><TD WIDTH=432 VALIGN=TOP><FONT FACE='Times New Roman, Times, Serif' SIZE=2>finds groups that maximize similarity within groups</FONT></TD></TR><TR><TD WIDTH=151 VALIGN=TOP><FONT FACE='Times New Roman, Times, Serif' SIZE=2><I>Exhaustive</I></FONT></TD><TD WIDTH=432 VALIGN=TOP><FONT FACE='Times New Roman, Times, Serif' SIZE=2>finds groups that maximize statistical significance</FONT></TD></TR></TABLE></TD></TR></TABLE><BR><A NAME='{5CE}'/></FORM></P></TD></TR></TABLE><P><FONT SIZE=0 COLOR=WHITE></CENTER><A NAME="bottom"> </A><!-- netLibrary.com Copyright Notice --> </BODY></HTML>⌨️ 快捷键说明
复制代码
Ctrl + C
搜索代码
Ctrl + F
全屏模式
F11
切换主题
Ctrl + Shift + D
显示快捷键
?
增大字号
Ctrl + =
减小字号
Ctrl + -