📄 page_169.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_168";var NextPage="Page_170";var CurPage="Page_169";var PageOrder="178";</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_168.html'>Previous</A></TD> <TD ALIGN=RIGHT><A HREF='Page_170.html'>Next</A></TD> </TR> </TABLE></TD></TR><TR><TD ALIGN=LEFT><P><A NAME='JUMPDEST_Page_169'/><A NAME='{5E4}'/><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 169</FONT></TD></TR></TABLE><A NAME='{5E5}'/><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>means for discovering effective network structure/input combinations. The one drawback to this approach is one of time. A complex analysis, say of 10 or 20 web pages on your site, will likely take hours or even days to complete. The other drawback of the NGO is that it can only construct models on less than 32,000 observations.</FONT></TD><TD></TD></TR><TR> <TD COLSPAN=3></TD></TR><TR> <TD COLSPAN=3 HEIGHT=1></TD></TR></TABLE><A NAME='{5E6}'/><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>This adaptive multimodeling capability is a good pragmatic approach to data mining highly dynamic data such as a web-generated database. The result of this process is higher accuracy neural networks that are optimized for the entire website with less effort on your part. The NGO automates much of the modeling development and data mining process. Some of the tedious tasks like the splitting of the testing/training data sets, determining which input variables to use and the overall neural network architectural design, are all automatically done for the user by the software. The NGO uses its genetic algorithms to evolve the neural network structures and select which input variables are the most important to what you are trying to predict, such as total sales for each product on your website. Once NGO has produced a model that fits the data, that is, learned the relationships between multiple inputs and a desired output, you need a separate product called the Predictor to view and use the resulting network. Predictor is shipped with the NGO and allows you to view graphs resulting from the network. It also allows you to enter data for predictions or load files to create scores that can be saved to disk.</FONT></TD><TD></TD></TR><TR> <TD COLSPAN=3></TD></TR><TR> <TD COLSPAN=3 HEIGHT=1></TD></TR></TABLE><A NAME='{5E7}'/><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>Using the NGO can be as simple as opening a data file and running the software. The tool prompts you for your data format type and loads it into memory. Next it prompts you for the identification of inputs and desired output. Since you are likely going to be dealing with an ASCII comma-delimited flat file, generated from your server, you must tell NGO which fields to use as inputs and what field will be your output. From that point NGO goes on to build and validates the training and testing data sets, providing a variety of methods for viewing the modeling processes. NGO provides multiple graphs for monitoring its modeling activity, such as <I>Predicted vs. Desired</I> or <I>Accuracy vs. Error Trend</I> results, <I>Status</I> and <I>Report</I> screens—all of them can be viewed from multiple screens. This is a good feature since once you start the mining process you need to know that the training and testing of the data are taking place in the background.</FONT></TD><TD></TD></TR><TR> <TD COLSPAN=3></TD></TR><TR> <TD COLSPAN=3 HEIGHT=1></TD></TR></TABLE><A NAME='{5E8}'/><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 NGO also creates a population of candidate-input variables and neural structures, training them and evaluating them. It next selects the top networks and pairs up the genetic material representing the inputs and neural structure of the networks and does the exchanges of the genetic material between them. As with all proper genetic algorithms it also throws in a few mutations for a flavor of</FONT><FONT FACE='Times New Roman, Times, Serif' SIZE=3 COLOR=#FFFF00><!-- continue --></FONT></TD><TD></TD></TR><TR> <TD COLSPAN=3></TD></TR><TR> <TD COLSPAN=3 HEIGHT=1></TD></TR></TABLE><A NAME='{5E9}'/></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 + -