2009-08-09 11:17:39 +02:00
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
< html > < head > < meta http-equiv = "Content-Type" content = "text/html;charset=UTF-8" >
< title > Neural++: neuralpp::NeuralNet Class Reference< / title >
< link href = "doxygen.css" rel = "stylesheet" type = "text/css" >
< link href = "tabs.css" rel = "stylesheet" type = "text/css" >
< / head > < body >
<!-- Generated by Doxygen 1.5.6 -->
< div class = "navigation" id = "top" >
< div class = "tabs" >
< ul >
< li > < a href = "index.html" > < span > Main Page< / span > < / a > < / li >
< li > < a href = "namespaces.html" > < span > Namespaces< / span > < / a > < / li >
< li class = "current" > < a href = "annotated.html" > < span > Classes< / span > < / a > < / li >
< li > < a href = "files.html" > < span > Files< / span > < / a > < / li >
< / ul >
< / div >
< div class = "tabs" >
< ul >
< li > < a href = "annotated.html" > < span > Class List< / span > < / a > < / li >
< li > < a href = "functions.html" > < span > Class Members< / span > < / a > < / li >
< / ul >
< / div >
< div class = "navpath" > < a class = "el" href = "namespaceneuralpp.html" > neuralpp< / a > ::< a class = "el" href = "classneuralpp_1_1NeuralNet.html" > NeuralNet< / a >
< / div >
< / div >
< div class = "contents" >
< h1 > neuralpp::NeuralNet Class Reference< / h1 > <!-- doxytag: class="neuralpp::NeuralNet" --> Main project's class.
< a href = "#_details" > More...< / a >
< p >
< code > #include < < a class = "el" href = "neural_09_09_8hpp-source.html" > neural++.hpp< / a > > < / code >
< p >
< p >
< a href = "classneuralpp_1_1NeuralNet-members.html" > List of all members.< / a > < table border = "0" cellpadding = "0" cellspacing = "0" >
< tr > < td > < / td > < / tr >
< tr > < td colspan = "2" > < br > < h2 > Public Types< / h2 > < / td > < / tr >
< tr > < td class = "memItemLeft" nowrap align = "right" valign = "top" > enum < / td > < td class = "memItemRight" valign = "bottom" > < a class = "el" href = "classneuralpp_1_1NeuralNet.html#94c36c94060e785ea67a0014c4182f8f" > source< / a > { < a class = "el" href = "classneuralpp_1_1NeuralNet.html#94c36c94060e785ea67a0014c4182f8f5ec2727c0756ddb097b53efe49b81afb" > file< / a > ,
< a class = "el" href = "classneuralpp_1_1NeuralNet.html#94c36c94060e785ea67a0014c4182f8f6d06b4fe9414a158c97aee1a3679a904" > str< / a >
}< / td > < / tr >
< tr > < td class = "mdescLeft" > < / td > < td class = "mdescRight" > Enum to choose the eventual training source for our network (XML from a file or from a string). < a href = "classneuralpp_1_1NeuralNet.html#94c36c94060e785ea67a0014c4182f8f" > More...< / a > < br > < / td > < / tr >
< tr > < td colspan = "2" > < br > < h2 > Public Member Functions< / h2 > < / td > < / tr >
< tr > < td class = "memItemLeft" nowrap align = "right" valign = "top" > < / td > < td class = "memItemRight" valign = "bottom" > < a class = "el" href = "classneuralpp_1_1NeuralNet.html#92b145f2f6f00bf1ba645ce2235882c2" > NeuralNet< / a > ()< / td > < / tr >
< tr > < td class = "mdescLeft" > < / td > < td class = "mdescRight" > Empty constructor for the class - it just makes nothing. < a href = "#92b145f2f6f00bf1ba645ce2235882c2" > < / a > < br > < / td > < / tr >
< tr > < td class = "memItemLeft" nowrap align = "right" valign = "top" > < / td > < td class = "memItemRight" valign = "bottom" > < a class = "el" href = "classneuralpp_1_1NeuralNet.html#c79534c7c0dfb20d1d03be2ad7569b78" > NeuralNet< / a > (size_t in_size, size_t hidden_size, size_t out_size, double l, int e)< / td > < / tr >
< tr > < td class = "mdescLeft" > < / td > < td class = "mdescRight" > Constructor. < a href = "#c79534c7c0dfb20d1d03be2ad7569b78" > < / a > < br > < / td > < / tr >
2009-08-15 02:59:09 +02:00
< tr > < td class = "memItemLeft" nowrap align = "right" valign = "top" > < / td > < td class = "memItemRight" valign = "bottom" > < a class = "el" href = "classneuralpp_1_1NeuralNet.html#7fc7fc3e3220c138ffa5356fef6b9757" > NeuralNet< / a > (const string file) throw (NetworkFileNotFoundException)< / td > < / tr >
2009-08-09 11:17:39 +02:00
2009-08-15 02:59:09 +02:00
< tr > < td class = "mdescLeft" > < / td > < td class = "mdescRight" > Constructor. < a href = "#7fc7fc3e3220c138ffa5356fef6b9757" > < / a > < br > < / td > < / tr >
< tr > < td class = "memItemLeft" nowrap align = "right" valign = "top" > < / td > < td class = "memItemRight" valign = "bottom" > < a class = "el" href = "classneuralpp_1_1NeuralNet.html#0c16df2e0701503052c63749930b238e" > NeuralNet< / a > (size_t in_size, size_t hidden_size, size_t out_size, double(*actv)(double), double l, int e)< / td > < / tr >
2009-08-09 11:17:39 +02:00
2009-08-15 02:59:09 +02:00
< tr > < td class = "mdescLeft" > < / td > < td class = "mdescRight" > Constructor. < a href = "#0c16df2e0701503052c63749930b238e" > < / a > < br > < / td > < / tr >
< tr > < td class = "memItemLeft" nowrap align = "right" valign = "top" > double < / td > < td class = "memItemRight" valign = "bottom" > < a class = "el" href = "classneuralpp_1_1NeuralNet.html#961dce8913264bf64c899dce4e25f810" > getOutput< / a > () const < / td > < / tr >
2009-08-09 11:17:39 +02:00
2009-08-15 02:59:09 +02:00
< tr > < td class = "mdescLeft" > < / td > < td class = "mdescRight" > It gets the output of the network (note: the layer output should contain an only neuron). < a href = "#961dce8913264bf64c899dce4e25f810" > < / a > < br > < / td > < / tr >
2009-08-09 11:17:39 +02:00
< tr > < td class = "memItemLeft" nowrap align = "right" valign = "top" > vector< double > < / td > < td class = "memItemRight" valign = "bottom" > < a class = "el" href = "classneuralpp_1_1NeuralNet.html#a6b8bf3800b43b58843c65fc431207ae" > getOutputs< / a > ()< / td > < / tr >
< tr > < td class = "mdescLeft" > < / td > < td class = "mdescRight" > It gets the output of the network in case the output layer contains more neurons. < a href = "#a6b8bf3800b43b58843c65fc431207ae" > < / a > < br > < / td > < / tr >
2009-08-15 02:59:09 +02:00
< tr > < td class = "memItemLeft" nowrap align = "right" valign = "top" > double < / td > < td class = "memItemRight" valign = "bottom" > < a class = "el" href = "classneuralpp_1_1NeuralNet.html#562dfe9fb8d73bf25a23ce608451d3aa" > expected< / a > () const < / td > < / tr >
2009-08-09 11:17:39 +02:00
2009-08-15 02:59:09 +02:00
< tr > < td class = "mdescLeft" > < / td > < td class = "mdescRight" > It gets the value expected. < a href = "#562dfe9fb8d73bf25a23ce608451d3aa" > < / a > < br > < / td > < / tr >
2009-08-09 11:17:39 +02:00
< tr > < td class = "memItemLeft" nowrap align = "right" valign = "top" > void < / td > < td class = "memItemRight" valign = "bottom" > < a class = "el" href = "classneuralpp_1_1NeuralNet.html#b6475762b7e9eab086befdc511f7c236" > setExpected< / a > (double < a class = "el" href = "classneuralpp_1_1NeuralNet.html#261f5f68fcc5be54250cfa03945266dd" > ex< / a > )< / td > < / tr >
< tr > < td class = "mdescLeft" > < / td > < td class = "mdescRight" > It sets the value you expect from your network. < a href = "#b6475762b7e9eab086befdc511f7c236" > < / a > < br > < / td > < / tr >
< tr > < td class = "memItemLeft" nowrap align = "right" valign = "top" > void < / td > < td class = "memItemRight" valign = "bottom" > < a class = "el" href = "classneuralpp_1_1NeuralNet.html#b0bd1daadb06980dff1f50d33a7c098e" > update< / a > ()< / td > < / tr >
< tr > < td class = "mdescLeft" > < / td > < td class = "mdescRight" > It updates through back-propagation the weights of the synapsis and computes again the output value for < em > epochs< / em > times, calling back updateWeights and commitChanges functions. < a href = "#b0bd1daadb06980dff1f50d33a7c098e" > < / a > < br > < / td > < / tr >
< tr > < td class = "memItemLeft" nowrap align = "right" valign = "top" > void < / td > < td class = "memItemRight" valign = "bottom" > < a class = "el" href = "classneuralpp_1_1NeuralNet.html#c129c180647362da963758bfd1ba6890" > propagate< / a > ()< / td > < / tr >
< tr > < td class = "mdescLeft" > < / td > < td class = "mdescRight" > It propagates values through the network. < a href = "#c129c180647362da963758bfd1ba6890" > < / a > < br > < / td > < / tr >
< tr > < td class = "memItemLeft" nowrap align = "right" valign = "top" > void < / td > < td class = "memItemRight" valign = "bottom" > < a class = "el" href = "classneuralpp_1_1NeuralNet.html#0de170e8ab561ad63d0739b4c4b74f68" > setInput< / a > (vector< double > & v)< / td > < / tr >
< tr > < td class = "mdescLeft" > < / td > < td class = "mdescRight" > It sets the input for the network. < a href = "#0de170e8ab561ad63d0739b4c4b74f68" > < / a > < br > < / td > < / tr >
< tr > < td class = "memItemLeft" nowrap align = "right" valign = "top" > void < / td > < td class = "memItemRight" valign = "bottom" > < a class = "el" href = "classneuralpp_1_1NeuralNet.html#46f23f462318a4ffc037a4e806364c3f" > link< / a > ()< / td > < / tr >
< tr > < td class = "mdescLeft" > < / td > < td class = "mdescRight" > It links the layers of the network (input, hidden, output). < a href = "#46f23f462318a4ffc037a4e806364c3f" > < / a > < br > < / td > < / tr >
2009-08-15 02:59:09 +02:00
< tr > < td class = "memItemLeft" nowrap align = "right" valign = "top" > void < / td > < td class = "memItemRight" valign = "bottom" > < a class = "el" href = "classneuralpp_1_1NeuralNet.html#fdf94c276720c25e565cac834fe8a407" > save< / a > (const char *fname) throw (NetworkFileWriteException)< / td > < / tr >
2009-08-09 11:17:39 +02:00
2009-08-15 02:59:09 +02:00
< tr > < td class = "mdescLeft" > < / td > < td class = "mdescRight" > Save a trained neural network to a binary file. < a href = "#fdf94c276720c25e565cac834fe8a407" > < / a > < br > < / td > < / tr >
< tr > < td class = "memItemLeft" nowrap align = "right" valign = "top" > void < / td > < td class = "memItemRight" valign = "bottom" > < a class = "el" href = "classneuralpp_1_1NeuralNet.html#ead4bdef0602a5cadbe3beb685e01f5f" > train< / a > (string xml, < a class = "el" href = "classneuralpp_1_1NeuralNet.html#94c36c94060e785ea67a0014c4182f8f" > source< / a > src) throw (InvalidXMLException)< / td > < / tr >
2009-08-09 11:17:39 +02:00
2009-08-15 02:59:09 +02:00
< tr > < td class = "mdescLeft" > < / td > < td class = "mdescRight" > Train a network using a training set loaded from an XML file. < a href = "#ead4bdef0602a5cadbe3beb685e01f5f" > < / a > < br > < / td > < / tr >
2009-08-09 11:17:39 +02:00
< tr > < td colspan = "2" > < br > < h2 > Static Public Member Functions< / h2 > < / td > < / tr >
< tr > < td class = "memItemLeft" nowrap align = "right" valign = "top" > static void < / td > < td class = "memItemRight" valign = "bottom" > < a class = "el" href = "classneuralpp_1_1NeuralNet.html#45c7645d4affe65752d37cd230afba24" > initXML< / a > (string & xml)< / td > < / tr >
< tr > < td class = "mdescLeft" > < / td > < td class = "mdescRight" > Initialize the training XML for the neural network. < a href = "#45c7645d4affe65752d37cd230afba24" > < / a > < br > < / td > < / tr >
< tr > < td class = "memItemLeft" nowrap align = "right" valign = "top" > static vector< double > < / td > < td class = "memItemRight" valign = "bottom" > < a class = "el" href = "classneuralpp_1_1NeuralNet.html#e07af23ceb8666518da0c035bf1e0376" > split< / a > (char delim, string str)< / td > < / tr >
< tr > < td class = "mdescLeft" > < / td > < td class = "mdescRight" > Splits a string into a vector of doubles, given a delimitator. < a href = "#e07af23ceb8666518da0c035bf1e0376" > < / a > < br > < / td > < / tr >
< tr > < td class = "memItemLeft" nowrap align = "right" valign = "top" > static string < / td > < td class = "memItemRight" valign = "bottom" > < a class = "el" href = "classneuralpp_1_1NeuralNet.html#4be31ecb0b543a192997bd83c6995ccb" > XMLFromSet< / a > (int id, string set)< / td > < / tr >
< tr > < td class = "mdescLeft" > < / td > < td class = "mdescRight" > Get a training set from a string and copies it to an XML For example, these strings could be training sets for making sums: "2,3;5" - "5,6;11" - "2,2;4" - "4,5:9" This method called on the first string will return an XML such this: '< training id="0"> < input id="0"> 2< /input> < input id="1"> 3< /input> < output id="0"> 5< /output> & lt/training> '. < a href = "#4be31ecb0b543a192997bd83c6995ccb" > < / a > < br > < / td > < / tr >
< tr > < td class = "memItemLeft" nowrap align = "right" valign = "top" > static void < / td > < td class = "memItemRight" valign = "bottom" > < a class = "el" href = "classneuralpp_1_1NeuralNet.html#28b9966c5f197b8e86d57dd104aa32a6" > closeXML< / a > (string & xml)< / td > < / tr >
< tr > < td class = "mdescLeft" > < / td > < td class = "mdescRight" > Closes an open XML document generated by "initXML" and "XMLFromSet". < a href = "#28b9966c5f197b8e86d57dd104aa32a6" > < / a > < br > < / td > < / tr >
2009-08-15 02:59:09 +02:00
< tr > < td colspan = "2" > < br > < h2 > Public Attributes< / h2 > < / td > < / tr >
< tr > < td class = "memItemLeft" nowrap align = "right" valign = "top" > < a class = "el" href = "classneuralpp_1_1Layer.html" > Layer< / a > * < / td > < td class = "memItemRight" valign = "bottom" > < a class = "el" href = "classneuralpp_1_1NeuralNet.html#e2b4e8405f9d25edab395d61502bdba9" > input< / a > < / td > < / tr >
< tr > < td class = "memItemLeft" nowrap align = "right" valign = "top" > < a class = "el" href = "classneuralpp_1_1Layer.html" > Layer< / a > * < / td > < td class = "memItemRight" valign = "bottom" > < a class = "el" href = "classneuralpp_1_1NeuralNet.html#bbdaa1b6c0a1a95d2b18cd25fda2a266" > hidden< / a > < / td > < / tr >
< tr > < td class = "memItemLeft" nowrap align = "right" valign = "top" > < a class = "el" href = "classneuralpp_1_1Layer.html" > Layer< / a > * < / td > < td class = "memItemRight" valign = "bottom" > < a class = "el" href = "classneuralpp_1_1NeuralNet.html#fa9b2dbcbb39d0fc70f790ac24069a74" > output< / a > < / td > < / tr >
2009-08-09 11:17:39 +02:00
< tr > < td colspan = "2" > < br > < h2 > Private Member Functions< / h2 > < / td > < / tr >
< tr > < td class = "memItemLeft" nowrap align = "right" valign = "top" > void < / td > < td class = "memItemRight" valign = "bottom" > < a class = "el" href = "classneuralpp_1_1NeuralNet.html#94169c89a7cd47122ab5dbf1d5c5e108" > updateWeights< / a > ()< / td > < / tr >
< tr > < td class = "mdescLeft" > < / td > < td class = "mdescRight" > It updates the weights of the net's synapsis through back-propagation. < a href = "#94169c89a7cd47122ab5dbf1d5c5e108" > < / a > < br > < / td > < / tr >
< tr > < td class = "memItemLeft" nowrap align = "right" valign = "top" > void < / td > < td class = "memItemRight" valign = "bottom" > < a class = "el" href = "classneuralpp_1_1NeuralNet.html#62695a82dfb1df758a44150921aec8e0" > commitChanges< / a > (< a class = "el" href = "classneuralpp_1_1Layer.html" > Layer< / a > *l)< / td > < / tr >
< tr > < td class = "mdescLeft" > < / td > < td class = "mdescRight" > It commits the changes made by < a class = "el" href = "classneuralpp_1_1NeuralNet.html#94169c89a7cd47122ab5dbf1d5c5e108" title = "It updates the weights of the net's synapsis through back-propagation." > updateWeights()< / a > to the layer l. < a href = "#62695a82dfb1df758a44150921aec8e0" > < / a > < br > < / td > < / tr >
2009-08-15 02:59:09 +02:00
< tr > < td class = "memItemLeft" nowrap align = "right" valign = "top" > double < / td > < td class = "memItemRight" valign = "bottom" > < a class = "el" href = "classneuralpp_1_1NeuralNet.html#0616c51404efaca2714e37dd7478997e" > error< / a > (double < a class = "el" href = "classneuralpp_1_1NeuralNet.html#261f5f68fcc5be54250cfa03945266dd" > ex< / a > ) const < / td > < / tr >
2009-08-09 11:17:39 +02:00
2009-08-15 02:59:09 +02:00
< tr > < td class = "mdescLeft" > < / td > < td class = "mdescRight" > Get the error made on the expected result as |v-v'|/v. < a href = "#0616c51404efaca2714e37dd7478997e" > < / a > < br > < / td > < / tr >
2009-08-09 11:17:39 +02:00
< tr > < td colspan = "2" > < br > < h2 > Private Attributes< / h2 > < / td > < / tr >
< tr > < td class = "memItemLeft" nowrap align = "right" valign = "top" > int < / td > < td class = "memItemRight" valign = "bottom" > < a class = "el" href = "classneuralpp_1_1NeuralNet.html#4cb52dae7b43d03fac73afca7b9f3a51" > epochs< / a > < / td > < / tr >
< tr > < td class = "memItemLeft" nowrap align = "right" valign = "top" > int < / td > < td class = "memItemRight" valign = "bottom" > < a class = "el" href = "classneuralpp_1_1NeuralNet.html#4f88106c9e542c39eac43b4ca1974a2a" > ref_epochs< / a > < / td > < / tr >
< tr > < td class = "memItemLeft" nowrap align = "right" valign = "top" > double < / td > < td class = "memItemRight" valign = "bottom" > < a class = "el" href = "classneuralpp_1_1NeuralNet.html#6bd7be443e46b2fdbf1da2edb8e611ab" > l_rate< / a > < / td > < / tr >
< tr > < td class = "memItemLeft" nowrap align = "right" valign = "top" > double < / td > < td class = "memItemRight" valign = "bottom" > < a class = "el" href = "classneuralpp_1_1NeuralNet.html#261f5f68fcc5be54250cfa03945266dd" > ex< / a > < / td > < / tr >
< tr > < td class = "memItemLeft" nowrap align = "right" valign = "top" > double(* < / td > < td class = "memItemRight" valign = "bottom" > < a class = "el" href = "classneuralpp_1_1NeuralNet.html#c1469e6afd87d85b82f14bc246f82457" > actv_f< / a > )(double)< / td > < / tr >
< tr > < td class = "mdescLeft" > < / td > < td class = "mdescRight" > Private pointer to function, containing the function to be used as activation function. < a href = "#c1469e6afd87d85b82f14bc246f82457" > < / a > < br > < / td > < / tr >
< / table >
< hr > < a name = "_details" > < / a > < h2 > Detailed Description< / h2 >
Main project's class.
< p >
Use *ONLY* this class, unless you know what you're doing < hr > < h2 > Member Enumeration Documentation< / h2 >
< a class = "anchor" name = "94c36c94060e785ea67a0014c4182f8f" > < / a > <!-- doxytag: member="neuralpp::NeuralNet::source" ref="94c36c94060e785ea67a0014c4182f8f" args="" -->
< div class = "memitem" >
< div class = "memproto" >
< table class = "memname" >
< tr >
< td class = "memname" > enum < a class = "el" href = "classneuralpp_1_1NeuralNet.html#94c36c94060e785ea67a0014c4182f8f" > neuralpp::NeuralNet::source< / a > < / td >
< / tr >
< / table >
< / div >
< div class = "memdoc" >
< p >
Enum to choose the eventual training source for our network (XML from a file or from a string).
< p >
< dl compact > < dt > < b > Enumerator: < / b > < / dt > < dd >
< table border = "0" cellspacing = "2" cellpadding = "0" >
< tr > < td valign = "top" > < em > < a class = "anchor" name = "94c36c94060e785ea67a0014c4182f8f5ec2727c0756ddb097b53efe49b81afb" > < / a > <!-- doxytag: member="file" ref="94c36c94060e785ea67a0014c4182f8f5ec2727c0756ddb097b53efe49b81afb" args="" --> file< / em > < / td > < td >
< / td > < / tr >
< tr > < td valign = "top" > < em > < a class = "anchor" name = "94c36c94060e785ea67a0014c4182f8f6d06b4fe9414a158c97aee1a3679a904" > < / a > <!-- doxytag: member="str" ref="94c36c94060e785ea67a0014c4182f8f6d06b4fe9414a158c97aee1a3679a904" args="" --> str< / em > < / td > < td >
< / td > < / tr >
< / table >
< / dl >
< / div >
< / div > < p >
< hr > < h2 > Constructor & Destructor Documentation< / h2 >
< a class = "anchor" name = "92b145f2f6f00bf1ba645ce2235882c2" > < / a > <!-- doxytag: member="neuralpp::NeuralNet::NeuralNet" ref="92b145f2f6f00bf1ba645ce2235882c2" args="()" -->
< div class = "memitem" >
< div class = "memproto" >
< table class = "memname" >
< tr >
< td class = "memname" > neuralpp::NeuralNet::NeuralNet < / td >
< td > (< / td >
< td class = "paramname" > < / td >
< td > ) < / td >
< td > < code > [inline]< / code > < / td >
< / tr >
< / table >
< / div >
< div class = "memdoc" >
< p >
Empty constructor for the class - it just makes nothing.
< p >
< / div >
< / div > < p >
< a class = "anchor" name = "c79534c7c0dfb20d1d03be2ad7569b78" > < / a > <!-- doxytag: member="neuralpp::NeuralNet::NeuralNet" ref="c79534c7c0dfb20d1d03be2ad7569b78" args="(size_t in_size, size_t hidden_size, size_t out_size, double l, int e)" -->
< div class = "memitem" >
< div class = "memproto" >
< table class = "memname" >
< tr >
< td class = "memname" > neuralpp::NeuralNet::NeuralNet < / td >
< td > (< / td >
< td class = "paramtype" > size_t < / td >
< td class = "paramname" > < em > in_size< / em > , < / td >
< / tr >
< tr >
< td class = "paramkey" > < / td >
< td > < / td >
< td class = "paramtype" > size_t < / td >
< td class = "paramname" > < em > hidden_size< / em > , < / td >
< / tr >
< tr >
< td class = "paramkey" > < / td >
< td > < / td >
< td class = "paramtype" > size_t < / td >
< td class = "paramname" > < em > out_size< / em > , < / td >
< / tr >
< tr >
< td class = "paramkey" > < / td >
< td > < / td >
< td class = "paramtype" > double < / td >
< td class = "paramname" > < em > l< / em > , < / td >
< / tr >
< tr >
< td class = "paramkey" > < / td >
< td > < / td >
< td class = "paramtype" > int < / td >
< td class = "paramname" > < em > e< / em > < / td > < td > < / td >
< / tr >
< tr >
< td > < / td >
< td > )< / td >
< td > < / td > < td > < / td > < td > < / td >
< / tr >
< / table >
< / div >
< div class = "memdoc" >
< p >
Constructor.
< p >
< dl compact > < dt > < b > Parameters:< / b > < / dt > < dd >
< table border = "0" cellspacing = "2" cellpadding = "0" >
< tr > < td valign = "top" > < / td > < td valign = "top" > < em > in_size< / em > < / td > < td > Size of the input layer < / td > < / tr >
< tr > < td valign = "top" > < / td > < td valign = "top" > < em > hidden_size< / em > < / td > < td > Size of the hidden layer < / td > < / tr >
< tr > < td valign = "top" > < / td > < td valign = "top" > < em > out_size< / em > < / td > < td > Size of the output layer < / td > < / tr >
< tr > < td valign = "top" > < / td > < td valign = "top" > < em > l< / em > < / td > < td > learn rate (get it after doing some experiments, but generally try to keep its value quite low to be more accurate) < / td > < / tr >
< tr > < td valign = "top" > < / td > < td valign = "top" > < em > e< / em > < / td > < td > Epochs (cycles) to execute (the most you execute, the most the network can be accurate for its purpose) < / td > < / tr >
< / table >
< / dl >
< / div >
< / div > < p >
2009-08-15 02:59:09 +02:00
< a class = "anchor" name = "7fc7fc3e3220c138ffa5356fef6b9757" > < / a > <!-- doxytag: member="neuralpp::NeuralNet::NeuralNet" ref="7fc7fc3e3220c138ffa5356fef6b9757" args="(const string file)" -->
2009-08-09 11:17:39 +02:00
< div class = "memitem" >
< div class = "memproto" >
< table class = "memname" >
< tr >
< td class = "memname" > neuralpp::NeuralNet::NeuralNet < / td >
< td > (< / td >
2009-08-15 02:59:09 +02:00
< td class = "paramtype" > const string < / td >
2009-08-09 11:17:39 +02:00
< td class = "paramname" > < em > file< / em > < / td >
< td > ) < / td >
< td > throw (NetworkFileNotFoundException)< / td >
< / tr >
< / table >
< / div >
< div class = "memdoc" >
< p >
Constructor.
< p >
< dl compact > < dt > < b > Parameters:< / b > < / dt > < dd >
< table border = "0" cellspacing = "2" cellpadding = "0" >
2009-08-15 02:59:09 +02:00
< tr > < td valign = "top" > < / td > < td valign = "top" > < em > file< / em > < / td > < td > Binary file containing a neural network previously saved by < a class = "el" href = "classneuralpp_1_1NeuralNet.html#fdf94c276720c25e565cac834fe8a407" title = "Save a trained neural network to a binary file." > save()< / a > method < / td > < / tr >
2009-08-09 11:17:39 +02:00
< / table >
< / dl >
< dl compact > < dt > < b > Exceptions:< / b > < / dt > < dd >
< table border = "0" cellspacing = "2" cellpadding = "0" >
< tr > < td valign = "top" > < / td > < td valign = "top" > < em > NetworkFileNotFoundException< / em > < / td > < td > < / td > < / tr >
< / table >
< / dl >
< / div >
< / div > < p >
2009-08-15 02:59:09 +02:00
< a class = "anchor" name = "0c16df2e0701503052c63749930b238e" > < / a > <!-- doxytag: member="neuralpp::NeuralNet::NeuralNet" ref="0c16df2e0701503052c63749930b238e" args="(size_t in_size, size_t hidden_size, size_t out_size, double(*actv)(double), double l, int e)" -->
2009-08-09 11:17:39 +02:00
< div class = "memitem" >
< div class = "memproto" >
< table class = "memname" >
< tr >
< td class = "memname" > neuralpp::NeuralNet::NeuralNet < / td >
< td > (< / td >
< td class = "paramtype" > size_t < / td >
< td class = "paramname" > < em > in_size< / em > , < / td >
< / tr >
< tr >
< td class = "paramkey" > < / td >
< td > < / td >
< td class = "paramtype" > size_t < / td >
< td class = "paramname" > < em > hidden_size< / em > , < / td >
< / tr >
< tr >
< td class = "paramkey" > < / td >
< td > < / td >
< td class = "paramtype" > size_t < / td >
< td class = "paramname" > < em > out_size< / em > , < / td >
< / tr >
< tr >
< td class = "paramkey" > < / td >
< td > < / td >
< td class = "paramtype" > double(*)(double) < / td >
< td class = "paramname" > < em > actv< / em > , < / td >
< / tr >
< tr >
< td class = "paramkey" > < / td >
< td > < / td >
< td class = "paramtype" > double < / td >
< td class = "paramname" > < em > l< / em > , < / td >
< / tr >
< tr >
< td class = "paramkey" > < / td >
< td > < / td >
< td class = "paramtype" > int < / td >
< td class = "paramname" > < em > e< / em > < / td > < td > < / td >
< / tr >
< tr >
< td > < / td >
< td > )< / td >
< td > < / td > < td > < / td > < td > < / td >
< / tr >
< / table >
< / div >
< div class = "memdoc" >
< p >
Constructor.
< p >
< dl compact > < dt > < b > Parameters:< / b > < / dt > < dd >
< table border = "0" cellspacing = "2" cellpadding = "0" >
< tr > < td valign = "top" > < / td > < td valign = "top" > < em > in_size< / em > < / td > < td > Size of the input layer < / td > < / tr >
< tr > < td valign = "top" > < / td > < td valign = "top" > < em > hidden_size< / em > < / td > < td > Size of the hidden layer < / td > < / tr >
< tr > < td valign = "top" > < / td > < td valign = "top" > < em > out_size< / em > < / td > < td > Size of the output layer < / td > < / tr >
< tr > < td valign = "top" > < / td > < td valign = "top" > < em > actv< / em > < / td > < td > Activation function to use (default: f(x)=x) < / td > < / tr >
< tr > < td valign = "top" > < / td > < td valign = "top" > < em > l< / em > < / td > < td > learn rate (get it after doing some experiments, but generally try to keep its value quite low to be more accurate) < / td > < / tr >
< tr > < td valign = "top" > < / td > < td valign = "top" > < em > e< / em > < / td > < td > Epochs (cycles) to execute (the most you execute, the most the network can be accurate for its purpose) < / td > < / tr >
< / table >
< / dl >
< / div >
< / div > < p >
< hr > < h2 > Member Function Documentation< / h2 >
< a class = "anchor" name = "94169c89a7cd47122ab5dbf1d5c5e108" > < / a > <!-- doxytag: member="neuralpp::NeuralNet::updateWeights" ref="94169c89a7cd47122ab5dbf1d5c5e108" args="()" -->
< div class = "memitem" >
< div class = "memproto" >
< table class = "memname" >
< tr >
< td class = "memname" > void neuralpp::NeuralNet::updateWeights < / td >
< td > (< / td >
< td class = "paramname" > < / td >
< td > ) < / td >
< td > < code > [private]< / code > < / td >
< / tr >
< / table >
< / div >
< div class = "memdoc" >
< p >
It updates the weights of the net's synapsis through back-propagation.
< p >
In-class use only
< / div >
< / div > < p >
< a class = "anchor" name = "62695a82dfb1df758a44150921aec8e0" > < / a > <!-- doxytag: member="neuralpp::NeuralNet::commitChanges" ref="62695a82dfb1df758a44150921aec8e0" args="(Layer *l)" -->
< div class = "memitem" >
< div class = "memproto" >
< table class = "memname" >
< tr >
< td class = "memname" > void neuralpp::NeuralNet::commitChanges < / td >
< td > (< / td >
< td class = "paramtype" > < a class = "el" href = "classneuralpp_1_1Layer.html" > Layer< / a > * < / td >
< td class = "paramname" > < em > l< / em > < / td >
< td > ) < / td >
< td > < code > [private]< / code > < / td >
< / tr >
< / table >
< / div >
< div class = "memdoc" >
< p >
It commits the changes made by < a class = "el" href = "classneuralpp_1_1NeuralNet.html#94169c89a7cd47122ab5dbf1d5c5e108" title = "It updates the weights of the net's synapsis through back-propagation." > updateWeights()< / a > to the layer l.
< p >
In-class use only < dl compact > < dt > < b > Parameters:< / b > < / dt > < dd >
< table border = "0" cellspacing = "2" cellpadding = "0" >
< tr > < td valign = "top" > < / td > < td valign = "top" > < em > l< / em > < / td > < td > < a class = "el" href = "classneuralpp_1_1Layer.html" title = "Class for managing layers of neurons." > Layer< / a > to commit the changes < / td > < / tr >
< / table >
< / dl >
< / div >
< / div > < p >
2009-08-15 02:59:09 +02:00
< a class = "anchor" name = "0616c51404efaca2714e37dd7478997e" > < / a > <!-- doxytag: member="neuralpp::NeuralNet::error" ref="0616c51404efaca2714e37dd7478997e" args="(double ex) const " -->
2009-08-09 11:17:39 +02:00
< div class = "memitem" >
< div class = "memproto" >
< table class = "memname" >
< tr >
< td class = "memname" > double neuralpp::NeuralNet::error < / td >
< td > (< / td >
< td class = "paramtype" > double < / td >
< td class = "paramname" > < em > ex< / em > < / td >
< td > ) < / td >
2009-08-15 02:59:09 +02:00
< td > const< code > [private]< / code > < / td >
2009-08-09 11:17:39 +02:00
< / tr >
< / table >
< / div >
< div class = "memdoc" >
< p >
2009-08-15 02:59:09 +02:00
Get the error made on the expected result as |v-v'|/v.
2009-08-09 11:17:39 +02:00
< p >
< dl compact > < dt > < b > Parameters:< / b > < / dt > < dd >
< table border = "0" cellspacing = "2" cellpadding = "0" >
< tr > < td valign = "top" > < / td > < td valign = "top" > < em > ex< / em > < / td > < td > Expected value < / td > < / tr >
< / table >
< / dl >
< dl class = "return" compact > < dt > < b > Returns:< / b > < / dt > < dd > Mean error < / dd > < / dl >
< / div >
< / div > < p >
2009-08-15 02:59:09 +02:00
< a class = "anchor" name = "961dce8913264bf64c899dce4e25f810" > < / a > <!-- doxytag: member="neuralpp::NeuralNet::getOutput" ref="961dce8913264bf64c899dce4e25f810" args="() const " -->
2009-08-09 11:17:39 +02:00
< div class = "memitem" >
< div class = "memproto" >
< table class = "memname" >
< tr >
< td class = "memname" > double neuralpp::NeuralNet::getOutput < / td >
< td > (< / td >
< td class = "paramname" > < / td >
< td > ) < / td >
2009-08-15 02:59:09 +02:00
< td > const< / td >
2009-08-09 11:17:39 +02:00
< / tr >
< / table >
< / div >
< div class = "memdoc" >
< p >
It gets the output of the network (note: the layer output should contain an only neuron).
< p >
< dl class = "return" compact > < dt > < b > Returns:< / b > < / dt > < dd > The output value of the network < / dd > < / dl >
< / div >
< / div > < p >
< a class = "anchor" name = "a6b8bf3800b43b58843c65fc431207ae" > < / a > <!-- doxytag: member="neuralpp::NeuralNet::getOutputs" ref="a6b8bf3800b43b58843c65fc431207ae" args="()" -->
< div class = "memitem" >
< div class = "memproto" >
< table class = "memname" >
< tr >
< td class = "memname" > vector< double> neuralpp::NeuralNet::getOutputs < / td >
< td > (< / td >
< td class = "paramname" > < / td >
< td > ) < / td >
< td > < / td >
< / tr >
< / table >
< / div >
< div class = "memdoc" >
< p >
It gets the output of the network in case the output layer contains more neurons.
< p >
< dl class = "return" compact > < dt > < b > Returns:< / b > < / dt > < dd > A vector containing the output values of the network < / dd > < / dl >
< / div >
< / div > < p >
2009-08-15 02:59:09 +02:00
< a class = "anchor" name = "562dfe9fb8d73bf25a23ce608451d3aa" > < / a > <!-- doxytag: member="neuralpp::NeuralNet::expected" ref="562dfe9fb8d73bf25a23ce608451d3aa" args="() const " -->
2009-08-09 11:17:39 +02:00
< div class = "memitem" >
< div class = "memproto" >
< table class = "memname" >
< tr >
< td class = "memname" > double neuralpp::NeuralNet::expected < / td >
< td > (< / td >
< td class = "paramname" > < / td >
< td > ) < / td >
2009-08-15 02:59:09 +02:00
< td > const< / td >
2009-08-09 11:17:39 +02:00
< / tr >
< / table >
< / div >
< div class = "memdoc" >
< p >
It gets the value expected.
< p >
Of course you should specify this when you build your network by using setExpected. < dl class = "return" compact > < dt > < b > Returns:< / b > < / dt > < dd > The expected output value for a certain training phase < / dd > < / dl >
< / div >
< / div > < p >
< a class = "anchor" name = "b6475762b7e9eab086befdc511f7c236" > < / a > <!-- doxytag: member="neuralpp::NeuralNet::setExpected" ref="b6475762b7e9eab086befdc511f7c236" args="(double ex)" -->
< div class = "memitem" >
< div class = "memproto" >
< table class = "memname" >
< tr >
< td class = "memname" > void neuralpp::NeuralNet::setExpected < / td >
< td > (< / td >
< td class = "paramtype" > double < / td >
< td class = "paramname" > < em > ex< / em > < / td >
< td > ) < / td >
< td > < / td >
< / tr >
< / table >
< / div >
< div class = "memdoc" >
< p >
It sets the value you expect from your network.
< p >
< dl compact > < dt > < b > Parameters:< / b > < / dt > < dd >
< table border = "0" cellspacing = "2" cellpadding = "0" >
< tr > < td valign = "top" > < / td > < td valign = "top" > < em > ex< / em > < / td > < td > Expected output value < / td > < / tr >
< / table >
< / dl >
< / div >
< / div > < p >
< a class = "anchor" name = "b0bd1daadb06980dff1f50d33a7c098e" > < / a > <!-- doxytag: member="neuralpp::NeuralNet::update" ref="b0bd1daadb06980dff1f50d33a7c098e" args="()" -->
< div class = "memitem" >
< div class = "memproto" >
< table class = "memname" >
< tr >
< td class = "memname" > void neuralpp::NeuralNet::update < / td >
< td > (< / td >
< td class = "paramname" > < / td >
< td > ) < / td >
< td > < / td >
< / tr >
< / table >
< / div >
< div class = "memdoc" >
< p >
It updates through back-propagation the weights of the synapsis and computes again the output value for < em > epochs< / em > times, calling back updateWeights and commitChanges functions.
< p >
< / div >
< / div > < p >
< a class = "anchor" name = "c129c180647362da963758bfd1ba6890" > < / a > <!-- doxytag: member="neuralpp::NeuralNet::propagate" ref="c129c180647362da963758bfd1ba6890" args="()" -->
< div class = "memitem" >
< div class = "memproto" >
< table class = "memname" >
< tr >
< td class = "memname" > void neuralpp::NeuralNet::propagate < / td >
< td > (< / td >
< td class = "paramname" > < / td >
< td > ) < / td >
< td > < / td >
< / tr >
< / table >
< / div >
< div class = "memdoc" >
< p >
It propagates values through the network.
< p >
Use this when you want to give an already trained network some new values the get to the output
< / div >
< / div > < p >
< a class = "anchor" name = "0de170e8ab561ad63d0739b4c4b74f68" > < / a > <!-- doxytag: member="neuralpp::NeuralNet::setInput" ref="0de170e8ab561ad63d0739b4c4b74f68" args="(vector< double > &v)" -->
< div class = "memitem" >
< div class = "memproto" >
< table class = "memname" >
< tr >
< td class = "memname" > void neuralpp::NeuralNet::setInput < / td >
< td > (< / td >
< td class = "paramtype" > vector< double > & < / td >
< td class = "paramname" > < em > v< / em > < / td >
< td > ) < / td >
< td > < / td >
< / tr >
< / table >
< / div >
< div class = "memdoc" >
< p >
It sets the input for the network.
< p >
< dl compact > < dt > < b > Parameters:< / b > < / dt > < dd >
< table border = "0" cellspacing = "2" cellpadding = "0" >
< tr > < td valign = "top" > < / td > < td valign = "top" > < em > v< / em > < / td > < td > Vector of doubles, containing the values to give to your network < / td > < / tr >
< / table >
< / dl >
< / div >
< / div > < p >
< a class = "anchor" name = "46f23f462318a4ffc037a4e806364c3f" > < / a > <!-- doxytag: member="neuralpp::NeuralNet::link" ref="46f23f462318a4ffc037a4e806364c3f" args="()" -->
< div class = "memitem" >
< div class = "memproto" >
< table class = "memname" >
< tr >
< td class = "memname" > void neuralpp::NeuralNet::link < / td >
< td > (< / td >
< td class = "paramname" > < / td >
< td > ) < / td >
< td > < / td >
< / tr >
< / table >
< / div >
< div class = "memdoc" >
< p >
It links the layers of the network (input, hidden, output).
< p >
Don't use unless you exactly know what you're doing, it is already called by the constructor
< / div >
< / div > < p >
2009-08-15 02:59:09 +02:00
< a class = "anchor" name = "fdf94c276720c25e565cac834fe8a407" > < / a > <!-- doxytag: member="neuralpp::NeuralNet::save" ref="fdf94c276720c25e565cac834fe8a407" args="(const char *fname)" -->
2009-08-09 11:17:39 +02:00
< div class = "memitem" >
< div class = "memproto" >
< table class = "memname" >
< tr >
2009-08-15 02:59:09 +02:00
< td class = "memname" > void neuralpp::NeuralNet::save < / td >
2009-08-09 11:17:39 +02:00
< td > (< / td >
< td class = "paramtype" > const char * < / td >
< td class = "paramname" > < em > fname< / em > < / td >
< td > ) < / td >
2009-08-15 02:59:09 +02:00
< td > throw (NetworkFileWriteException)< / td >
2009-08-09 11:17:39 +02:00
< / tr >
< / table >
< / div >
< div class = "memdoc" >
< p >
Save a trained neural network to a binary file.
< p >
< dl compact > < dt > < b > Parameters:< / b > < / dt > < dd >
< table border = "0" cellspacing = "2" cellpadding = "0" >
< tr > < td valign = "top" > < / td > < td valign = "top" > < em > fname< / em > < / td > < td > Binary file where you're going to save your network < / td > < / tr >
< / table >
< / dl >
2009-08-15 02:59:09 +02:00
< dl compact > < dt > < b > Exceptions:< / b > < / dt > < dd >
< table border = "0" cellspacing = "2" cellpadding = "0" >
< tr > < td valign = "top" > < / td > < td valign = "top" > < em > NetworkFileWriteException< / em > < / td > < td > When you get an error writing the network's information to a file < / td > < / tr >
< / table >
< / dl >
2009-08-09 11:17:39 +02:00
< / div >
< / div > < p >
2009-08-15 02:59:09 +02:00
< a class = "anchor" name = "ead4bdef0602a5cadbe3beb685e01f5f" > < / a > <!-- doxytag: member="neuralpp::NeuralNet::train" ref="ead4bdef0602a5cadbe3beb685e01f5f" args="(string xml, source src)" -->
2009-08-09 11:17:39 +02:00
< div class = "memitem" >
< div class = "memproto" >
< table class = "memname" >
< tr >
< td class = "memname" > void neuralpp::NeuralNet::train < / td >
< td > (< / td >
< td class = "paramtype" > string < / td >
< td class = "paramname" > < em > xml< / em > , < / td >
< / tr >
< tr >
< td class = "paramkey" > < / td >
< td > < / td >
< td class = "paramtype" > < a class = "el" href = "classneuralpp_1_1NeuralNet.html#94c36c94060e785ea67a0014c4182f8f" > source< / a > < / td >
2009-08-15 02:59:09 +02:00
< td class = "paramname" > < em > src< / em > < / td > < td > < / td >
2009-08-09 11:17:39 +02:00
< / tr >
< tr >
< td > < / td >
< td > )< / td >
< td > < / td > < td > < / td > < td > throw (InvalidXMLException)< / td >
< / tr >
< / table >
< / div >
< div class = "memdoc" >
< p >
Train a network using a training set loaded from an XML file.
< p >
A sample XML file is available in examples/adder.xml < dl compact > < dt > < b > Parameters:< / b > < / dt > < dd >
< table border = "0" cellspacing = "2" cellpadding = "0" >
< tr > < td valign = "top" > < / td > < td valign = "top" > < em > xml< / em > < / td > < td > XML file containing our training set < / td > < / tr >
< tr > < td valign = "top" > < / td > < td valign = "top" > < em > src< / em > < / td > < td > Source type from which the XML will be loaded (from a file [default] or from a string) < / td > < / tr >
< / table >
< / dl >
< dl compact > < dt > < b > Exceptions:< / b > < / dt > < dd >
< table border = "0" cellspacing = "2" cellpadding = "0" >
< tr > < td valign = "top" > < / td > < td valign = "top" > < em > InvalidXMLException< / em > < / td > < td > < / td > < / tr >
< / table >
< / dl >
< / div >
< / div > < p >
< a class = "anchor" name = "45c7645d4affe65752d37cd230afba24" > < / a > <!-- doxytag: member="neuralpp::NeuralNet::initXML" ref="45c7645d4affe65752d37cd230afba24" args="(string &xml)" -->
< div class = "memitem" >
< div class = "memproto" >
< table class = "memname" >
< tr >
< td class = "memname" > static void neuralpp::NeuralNet::initXML < / td >
< td > (< / td >
< td class = "paramtype" > string & < / td >
< td class = "paramname" > < em > xml< / em > < / td >
< td > ) < / td >
< td > < code > [static]< / code > < / td >
< / tr >
< / table >
< / div >
< div class = "memdoc" >
< p >
Initialize the training XML for the neural network.
< p >
< dl compact > < dt > < b > Parameters:< / b > < / dt > < dd >
< table border = "0" cellspacing = "2" cellpadding = "0" >
< tr > < td valign = "top" > < / td > < td valign = "top" > < em > xml< / em > < / td > < td > String that will contain the XML < / td > < / tr >
< / table >
< / dl >
< / div >
< / div > < p >
< a class = "anchor" name = "e07af23ceb8666518da0c035bf1e0376" > < / a > <!-- doxytag: member="neuralpp::NeuralNet::split" ref="e07af23ceb8666518da0c035bf1e0376" args="(char delim, string str)" -->
< div class = "memitem" >
< div class = "memproto" >
< table class = "memname" >
< tr >
< td class = "memname" > static vector< double> neuralpp::NeuralNet::split < / td >
< td > (< / td >
< td class = "paramtype" > char < / td >
< td class = "paramname" > < em > delim< / em > , < / td >
< / tr >
< tr >
< td class = "paramkey" > < / td >
< td > < / td >
< td class = "paramtype" > string < / td >
< td class = "paramname" > < em > str< / em > < / td > < td > < / td >
< / tr >
< tr >
< td > < / td >
< td > )< / td >
< td > < / td > < td > < / td > < td > < code > [static]< / code > < / td >
< / tr >
< / table >
< / div >
< div class = "memdoc" >
< p >
Splits a string into a vector of doubles, given a delimitator.
< p >
< dl compact > < dt > < b > Parameters:< / b > < / dt > < dd >
< table border = "0" cellspacing = "2" cellpadding = "0" >
< tr > < td valign = "top" > < / td > < td valign = "top" > < em > delim< / em > < / td > < td > Delimitator < / td > < / tr >
< tr > < td valign = "top" > < / td > < td valign = "top" > < em > str< / em > < / td > < td > String to be splitted < / td > < / tr >
< / table >
< / dl >
< dl class = "return" compact > < dt > < b > Returns:< / b > < / dt > < dd > Vector of doubles containing splitted values < / dd > < / dl >
< / div >
< / div > < p >
< a class = "anchor" name = "4be31ecb0b543a192997bd83c6995ccb" > < / a > <!-- doxytag: member="neuralpp::NeuralNet::XMLFromSet" ref="4be31ecb0b543a192997bd83c6995ccb" args="(int id, string set)" -->
< div class = "memitem" >
< div class = "memproto" >
< table class = "memname" >
< tr >
< td class = "memname" > static string neuralpp::NeuralNet::XMLFromSet < / td >
< td > (< / td >
< td class = "paramtype" > int < / td >
< td class = "paramname" > < em > id< / em > , < / td >
< / tr >
< tr >
< td class = "paramkey" > < / td >
< td > < / td >
< td class = "paramtype" > string < / td >
< td class = "paramname" > < em > set< / em > < / td > < td > < / td >
< / tr >
< tr >
< td > < / td >
< td > )< / td >
< td > < / td > < td > < / td > < td > < code > [static]< / code > < / td >
< / tr >
< / table >
< / div >
< div class = "memdoc" >
< p >
Get a training set from a string and copies it to an XML For example, these strings could be training sets for making sums: "2,3;5" - "5,6;11" - "2,2;4" - "4,5:9" This method called on the first string will return an XML such this: '< training id="0"> < input id="0"> 2< /input> < input id="1"> 3< /input> < output id="0"> 5< /output> & lt/training> '.
< p >
< dl compact > < dt > < b > Parameters:< / b > < / dt > < dd >
< table border = "0" cellspacing = "2" cellpadding = "0" >
< tr > < td valign = "top" > < / td > < td valign = "top" > < em > id< / em > < / td > < td > ID for the given training set (0,1,..,n) < / td > < / tr >
< tr > < td valign = "top" > < / td > < td valign = "top" > < em > set< / em > < / td > < td > String containing input values and expected outputs < / td > < / tr >
< / table >
< / dl >
< dl class = "return" compact > < dt > < b > Returns:< / b > < / dt > < dd > XML string < / dd > < / dl >
< / div >
< / div > < p >
< a class = "anchor" name = "28b9966c5f197b8e86d57dd104aa32a6" > < / a > <!-- doxytag: member="neuralpp::NeuralNet::closeXML" ref="28b9966c5f197b8e86d57dd104aa32a6" args="(string &xml)" -->
< div class = "memitem" >
< div class = "memproto" >
< table class = "memname" >
< tr >
< td class = "memname" > static void neuralpp::NeuralNet::closeXML < / td >
< td > (< / td >
< td class = "paramtype" > string & < / td >
< td class = "paramname" > < em > xml< / em > < / td >
< td > ) < / td >
< td > < code > [static]< / code > < / td >
< / tr >
< / table >
< / div >
< div class = "memdoc" >
< p >
Closes an open XML document generated by "initXML" and "XMLFromSet".
< p >
< dl compact > < dt > < b > Parameters:< / b > < / dt > < dd >
< table border = "0" cellspacing = "2" cellpadding = "0" >
< tr > < td valign = "top" > < / td > < td valign = "top" > < em > xml< / em > < / td > < td > XML string to be closed < / td > < / tr >
< / table >
< / dl >
< / div >
< / div > < p >
< hr > < h2 > Member Data Documentation< / h2 >
< a class = "anchor" name = "4cb52dae7b43d03fac73afca7b9f3a51" > < / a > <!-- doxytag: member="neuralpp::NeuralNet::epochs" ref="4cb52dae7b43d03fac73afca7b9f3a51" args="" -->
< div class = "memitem" >
< div class = "memproto" >
< table class = "memname" >
< tr >
< td class = "memname" > int < a class = "el" href = "classneuralpp_1_1NeuralNet.html#4cb52dae7b43d03fac73afca7b9f3a51" > neuralpp::NeuralNet::epochs< / a > < code > [private]< / code > < / td >
< / tr >
< / table >
< / div >
< div class = "memdoc" >
< p >
< / div >
< / div > < p >
< a class = "anchor" name = "4f88106c9e542c39eac43b4ca1974a2a" > < / a > <!-- doxytag: member="neuralpp::NeuralNet::ref_epochs" ref="4f88106c9e542c39eac43b4ca1974a2a" args="" -->
< div class = "memitem" >
< div class = "memproto" >
< table class = "memname" >
< tr >
< td class = "memname" > int < a class = "el" href = "classneuralpp_1_1NeuralNet.html#4f88106c9e542c39eac43b4ca1974a2a" > neuralpp::NeuralNet::ref_epochs< / a > < code > [private]< / code > < / td >
< / tr >
< / table >
< / div >
< div class = "memdoc" >
< p >
< / div >
< / div > < p >
< a class = "anchor" name = "6bd7be443e46b2fdbf1da2edb8e611ab" > < / a > <!-- doxytag: member="neuralpp::NeuralNet::l_rate" ref="6bd7be443e46b2fdbf1da2edb8e611ab" args="" -->
< div class = "memitem" >
< div class = "memproto" >
< table class = "memname" >
< tr >
< td class = "memname" > double < a class = "el" href = "classneuralpp_1_1NeuralNet.html#6bd7be443e46b2fdbf1da2edb8e611ab" > neuralpp::NeuralNet::l_rate< / a > < code > [private]< / code > < / td >
< / tr >
< / table >
< / div >
< div class = "memdoc" >
< p >
< / div >
< / div > < p >
< a class = "anchor" name = "261f5f68fcc5be54250cfa03945266dd" > < / a > <!-- doxytag: member="neuralpp::NeuralNet::ex" ref="261f5f68fcc5be54250cfa03945266dd" args="" -->
< div class = "memitem" >
< div class = "memproto" >
< table class = "memname" >
< tr >
< td class = "memname" > double < a class = "el" href = "classneuralpp_1_1NeuralNet.html#261f5f68fcc5be54250cfa03945266dd" > neuralpp::NeuralNet::ex< / a > < code > [private]< / code > < / td >
< / tr >
< / table >
< / div >
< div class = "memdoc" >
< p >
< / div >
< / div > < p >
2009-08-15 02:59:09 +02:00
< a class = "anchor" name = "c1469e6afd87d85b82f14bc246f82457" > < / a > <!-- doxytag: member="neuralpp::NeuralNet::actv_f" ref="c1469e6afd87d85b82f14bc246f82457" args=")(double)" -->
2009-08-09 11:17:39 +02:00
< div class = "memitem" >
< div class = "memproto" >
< table class = "memname" >
< tr >
2009-08-15 02:59:09 +02:00
< td class = "memname" > double(* < a class = "el" href = "classneuralpp_1_1NeuralNet.html#c1469e6afd87d85b82f14bc246f82457" > neuralpp::NeuralNet::actv_f< / a > )(double)< code > [private]< / code > < / td >
2009-08-09 11:17:39 +02:00
< / tr >
< / table >
< / div >
< div class = "memdoc" >
< p >
2009-08-15 02:59:09 +02:00
Private pointer to function, containing the function to be used as activation function.
2009-08-09 11:17:39 +02:00
< p >
< / div >
< / div > < p >
2009-08-15 02:59:09 +02:00
< a class = "anchor" name = "e2b4e8405f9d25edab395d61502bdba9" > < / a > <!-- doxytag: member="neuralpp::NeuralNet::input" ref="e2b4e8405f9d25edab395d61502bdba9" args="" -->
2009-08-09 11:17:39 +02:00
< div class = "memitem" >
< div class = "memproto" >
< table class = "memname" >
< tr >
2009-08-15 02:59:09 +02:00
< td class = "memname" > < a class = "el" href = "classneuralpp_1_1Layer.html" > Layer< / a > * < a class = "el" href = "classneuralpp_1_1NeuralNet.html#e2b4e8405f9d25edab395d61502bdba9" > neuralpp::NeuralNet::input< / a > < / td >
2009-08-09 11:17:39 +02:00
< / tr >
< / table >
< / div >
< div class = "memdoc" >
< p >
< / div >
< / div > < p >
2009-08-15 02:59:09 +02:00
< a class = "anchor" name = "bbdaa1b6c0a1a95d2b18cd25fda2a266" > < / a > <!-- doxytag: member="neuralpp::NeuralNet::hidden" ref="bbdaa1b6c0a1a95d2b18cd25fda2a266" args="" -->
2009-08-09 11:17:39 +02:00
< div class = "memitem" >
< div class = "memproto" >
< table class = "memname" >
< tr >
2009-08-15 02:59:09 +02:00
< td class = "memname" > < a class = "el" href = "classneuralpp_1_1Layer.html" > Layer< / a > * < a class = "el" href = "classneuralpp_1_1NeuralNet.html#bbdaa1b6c0a1a95d2b18cd25fda2a266" > neuralpp::NeuralNet::hidden< / a > < / td >
2009-08-09 11:17:39 +02:00
< / tr >
< / table >
< / div >
< div class = "memdoc" >
< p >
< / div >
< / div > < p >
2009-08-15 02:59:09 +02:00
< a class = "anchor" name = "fa9b2dbcbb39d0fc70f790ac24069a74" > < / a > <!-- doxytag: member="neuralpp::NeuralNet::output" ref="fa9b2dbcbb39d0fc70f790ac24069a74" args="" -->
2009-08-09 11:17:39 +02:00
< div class = "memitem" >
< div class = "memproto" >
< table class = "memname" >
< tr >
2009-08-15 02:59:09 +02:00
< td class = "memname" > < a class = "el" href = "classneuralpp_1_1Layer.html" > Layer< / a > * < a class = "el" href = "classneuralpp_1_1NeuralNet.html#fa9b2dbcbb39d0fc70f790ac24069a74" > neuralpp::NeuralNet::output< / a > < / td >
2009-08-09 11:17:39 +02:00
< / tr >
< / table >
< / div >
< div class = "memdoc" >
< p >
< / div >
< / div > < p >
< hr > The documentation for this class was generated from the following file:< ul >
< li > < a class = "el" href = "neural_09_09_8hpp-source.html" > neural++.hpp< / a > < / ul >
< / div >
2009-08-15 02:59:09 +02:00
< hr size = "1" > < address style = "text-align: right;" > < small > Generated on Sat Aug 15 02:56:02 2009 for Neural++ by
2009-08-09 11:17:39 +02:00
< a href = "http://www.doxygen.org/index.html" >
< img src = "doxygen.png" alt = "doxygen" align = "middle" border = "0" > < / a > 1.5.6 < / small > < / address >
< / body >
< / html >
