📄 page_95.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_94";var NextPage="Page_96";var CurPage="Page_95";var PageOrder="106";</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_94.html'>Previous</A></TD> <TD ALIGN=RIGHT><A HREF='Page_96.html'>Next</A></TD> </TR> </TABLE></TD></TR><TR><TD ALIGN=LEFT><P><A NAME='JUMPDEST_Page_95'/><A NAME='{341}'/><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 95</FONT></TD></TR></TABLE><A NAME='{342}'/><TABLE BORDER=0 CELLSPACING=0 CELLPADDING=0><TR> <TD ROWSPAN=5></TD> <TD COLSPAN=3 HEIGHT=17></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><B>Training Cycle</B></FONT></TD><TD></TD></TR><TR> <TD COLSPAN=3></TD></TR><TR> <TD COLSPAN=3 HEIGHT=1></TD></TR></TABLE><A NAME='{343}'/><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>Your main task is ensuring you are capturing the relevant and important features from your website, which can be used for prediction and classification. As such, your objectives when using a neural network tool are to ensure you:</FONT></TD><TD></TD></TR><TR> <TD COLSPAN=3></TD></TR><TR> <TD COLSPAN=3 HEIGHT=1></TD></TR></TABLE><A NAME='{344}'/><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>1. Identify the input variables—this is <I>very</I> important!</FONT></TD><TD></TD></TR><TR> <TD COLSPAN=3></TD></TR><TR> <TD COLSPAN=3 HEIGHT=1></TD></TR></TABLE><A NAME='{345}'/><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>2. Convert the variables into usable ranges—pick a tool which will do this.</FONT></TD><TD></TD></TR><TR> <TD COLSPAN=3></TD></TR><TR> <TD COLSPAN=3 HEIGHT=1></TD></TR></TABLE><A NAME='{346}'/><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>3. Decide on the format of the output—continuous or categorical?</FONT></TD><TD></TD></TR><TR> <TD COLSPAN=3></TD></TR><TR> <TD COLSPAN=3 HEIGHT=1></TD></TR></TABLE><A NAME='{347}'/><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>4. Decide on the training data set sample and a training schedule.</FONT></TD><TD></TD></TR><TR> <TD COLSPAN=3></TD></TR><TR> <TD COLSPAN=3 HEIGHT=1></TD></TR></TABLE><A NAME='{348}'/><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>5. Test the model and apply it to your website objectives.</FONT></TD><TD></TD></TR><TR> <TD COLSPAN=3></TD></TR><TR> <TD COLSPAN=3 HEIGHT=1></TD></TR></TABLE><A NAME='{349}'/><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>Setting up a schedule for the training of a neural network is important because your model will deteriorate gradually over time. One of the downfalls of any model, especially one designed to predict web behavior, is that it will eventually go stale. The model you create with a neural network will age and atrophy as your website traffic evolves beyond the data used to train your net. As you introduce new products and services, or as you make changes and improvements to your website, changes occur in your traffic and transaction patterns. As such you will need to train your network with fresh data.</FONT></TD><TD></TD></TR><TR> <TD COLSPAN=3></TD></TR><TR> <TD COLSPAN=3 HEIGHT=1></TD></TR></TABLE><A NAME='{34A}'/><TABLE BORDER=0 CELLSPACING=0 CELLPADDING=0><TR> <TD ROWSPAN=5></TD> <TD COLSPAN=3 HEIGHT=17></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><B>Self-Organizing Maps (SOM)</B></FONT></TD><TD></TD></TR><TR> <TD COLSPAN=3></TD></TR><TR> <TD COLSPAN=3 HEIGHT=1></TD></TR></TABLE><A NAME='{34B}'/><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>Neural networks perform best in situations whose outputs are nonlinear and with complex problems that do not require a solution of absolute accuracy. They also are ideal in situations where a large amount of historical data is available for training and where a "black box" solution will do and no explanation is required. Most of today's data mining tools incorporate two main types of neural networks, back-propagation and SOMs, also known as Kohonen networks, after its inventor.</FONT></TD><TD></TD></TR><TR> <TD COLSPAN=3></TD></TR><TR> <TD COLSPAN=3 HEIGHT=1></TD></TR></TABLE><A NAME='{34C}'/><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>In 1981 Tuevo Kohonen proposed and demonstrated a completely different network architecture. An SOM network resembles a collection of neurons, as in the human brain, each connected to its neighbor. As an SOM is exposed to samples, it begins to create self-organizing clusters, like cellophane spreading itself over chunks of data. Gradually, a Kohonen network will organize clusters and can thus be used for situations where no output or dependent variable is known. SOMs have been used to discover associations for such purposes as market basket analysis in retailing.</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='{34D}'/></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 + -