<snapdata remixID="14494599"><project name="Artificial Neural Networks - Sprite Layers But better" app="Snap! 11.0.0, https://snap.berkeley.edu" version="2"><notes>Building kit for artificial neural networks. Use a single "Layer" sprite to make a classical Rosenblatt perceptron. Duplicate the Layer sprite - several times - and adjust the receivers in the broadcast blocks to create deep neural networks. User the setup script to customize the network&apos;s topology and learning rate.</notes><thumbnail>data:image/png;base64,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</thumbnail><scenes select="1"><scene name="Artificial Neural Networks - Sprite Layers But better"><notes>Building kit for artificial neural networks. Use a single "Layer" sprite to make a classical Rosenblatt perceptron. Duplicate the Layer sprite - several times - and adjust the receivers in the broadcast blocks to create deep neural networks. User the setup script to customize the network&apos;s topology and learning rate.</notes><hidden></hidden><headers></headers><code></code><blocks><block-definition s="plot bars %&apos;data&apos; fill %&apos;width&apos; center %&apos;switch&apos;" type="command" category="pen"><comment x="0" y="0" w="120" collapsed="false">draw a list of numbers as  vertical lines distributed evenly across the stage.</comment><header></header><code></code><translations>de:male Balken _ gefüllt _ zentriert _&#xD;</translations><inputs><input type="%l"></input><input type="%n">0.8<options>single=0.8&#xD;pan=1&#xD;overlap=1.2</options></input><input type="%b">false</input></inputs><script><block s="clear"></block><block s="doDeclareVariables"><list><l>slice</l><l>pos</l><l>pen size</l></list></block><block s="doSetVar"><l>pos</l><block s="getPosition"></block></block><block s="doSetVar"><l>slice</l><block s="reportQuotient"><block s="reportAttributeOf"><l><option>width</option></l><block s="reportGet"><l><option>stage</option></l></block></block><block s="reportListAttribute"><l><option>length</option></l><block var="data"/></block></block></block><block s="doSetVar"><l>pen size</l><block s="getPenAttribute"><l><option>size</option></l></block></block><block s="setSize"><block s="reportVariadicProduct"><list><block var="slice"/><block var="width"/></list></block></block><block s="setXPosition"><block s="reportVariadicSum"><list><block s="reportAttributeOf"><l><option>left</option></l><block s="reportGet"><l><option>stage</option></l></block></block><block s="reportQuotient"><block var="slice"/><l>2</l></block></list></block></block><block s="doWarp"><script><block s="doForEach"><l>item</l><block var="data"/><script><block s="setYPosition"><block s="reportIfElse"><block var="switch"/><block s="reportQuotient"><block var="item"/><l>-2</l></block><block s="reportAttributeOf"><l><option>bottom</option></l><block s="reportGet"><l><option>stage</option></l></block></block></block></block><block s="down"></block><block s="changeYPosition"><block var="item"/></block><block s="up"></block><block s="changeXPosition"><block var="slice"/></block></script></block></script></block><block s="doGotoObject"><block var="pos"/></block><block s="setSize"><block var="pen size"/></block></script></block-definition></blocks><primitives></primitives><stage name="Stage" width="480" height="360" costume="0" color="255,255,255,1" tempo="60" threadsafe="false" penlog="false" volume="100" pan="0" lines="round" ternary="false" hyperops="true" codify="false" inheritance="true" sublistIDs="false" id="92"><pentrails>data:image/png;base64,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</pentrails><costumes><list struct="atomic" id="93"></list></costumes><sounds><list struct="atomic" id="94"></list></sounds><variables></variables><blocks></blocks><scripts></scripts><sprites select="1"><sprite name="ANN" idx="1" x="-150.00000000516798" y="-137.99999998865917" heading="90" scale="1" volume="100" pan="0" rotation="1" draggable="true" hidden="true" costume="0" color="80,80,80,1" pen="tip" id="99"><costumes><list struct="atomic" id="100"></list></costumes><sounds><list struct="atomic" id="101"></list></sounds><blocks></blocks><variables><variable name="target"><l>1</l></variable><variable name="log" transient="true"/><variable name="errors"><l>3.0007370591156164</l></variable></variables><scripts><comment x="20" y="10" w="399" collapsed="false">Sample Data with &quot;Rules&quot;:&#xD;the first 2 columns represent the data&apos;s &quot;features&quot;, the 3rd column is the expected result.</comment><script x="20" y="75.99999999999989"><block s="reportNewList"><list><block s="reportNewList"><list><l>0</l><l>0</l><l>0</l></list></block><block s="reportNewList"><list><l>1</l><l>0</l><l>0</l></list></block><block s="reportNewList"><list><l>0</l><l>1</l><l>0</l></list></block><block s="reportNewList"><list><l>1</l><l>1</l><l>1</l></list></block></list><comment w="90" collapsed="true">&quot;all&quot; -- AND</comment></block></script><script x="20" y="112.99999999999997"><block s="reportNewList"><list><block s="reportNewList"><list><l>0</l><l>0</l><l>1</l></list></block><block s="reportNewList"><list><l>1</l><l>0</l><l>0</l></list></block><block s="reportNewList"><list><l>0</l><l>1</l><l>0</l></list></block><block s="reportNewList"><list><l>1</l><l>1</l><l>1</l></list></block></list><comment w="116" collapsed="true">&quot;all er nuthin&quot; - XNOR</comment></block></script><script x="20" y="149.99999999999952"><block s="reportNewList"><list><block s="reportNewList"><list><l>0</l><l>0</l><l>0</l></list></block><block s="reportNewList"><list><l>1</l><l>0</l><l>1</l></list></block><block s="reportNewList"><list><l>0</l><l>1</l><l>1</l></list></block><block s="reportNewList"><list><l>1</l><l>1</l><l>1</l></list></block></list><comment w="90" collapsed="true">&quot;some&quot; - OR</comment></block></script><script x="20" y="186.99999999999943"><block s="doSetVar"><l>data</l><block s="reportNewList"><list><block s="reportNewList"><list><l>0</l><l>0</l><l>0</l></list></block><block s="reportNewList"><list><l>1</l><l>0</l><l>1</l></list></block><block s="reportNewList"><list><l>0</l><l>1</l><l>1</l></list></block><block s="reportNewList"><list><l>1</l><l>1</l><l>0</l></list></block></list><comment w="90" collapsed="true">&quot;either&quot; - XOR</comment></block></block></script><script x="20" y="654.999999999999"><block s="receiveMessage"><l>forward</l><list><l>output</l></list></block><block s="doDeclareVariables"><list><l>error</l></list></block><block s="doSetVar"><l>error</l><block s="reportDifference"><block var="target"/><block var="output"/></block></block><block s="doChangeVar"><l>errors</l><block s="reportVariadicSum"><block s="reportMonadic"><l><option>abs</option></l><block var="error"/></block></block></block><block s="doBroadcastAndWait"><l>backpropagate</l><list><l>Layer2</l><block var="error"/></list></block></script><script x="161" y="243.99999999999977"><block s="receiveGo"></block><block s="doDeclareVariables"><list><l>setup</l></list></block><block s="doForever"><script><block s="doSetVar"><l>setup</l><block s="reportNewList"><list><l>2</l><l>2</l><l>1</l></list></block><comment w="259.4826171874997" collapsed="true">layer sizes - # of neurons not counting bias</comment></block><block s="doFor"><l>i</l><l>1</l><block s="reportDifference"><block s="reportListAttribute"><l><option>length</option></l><block var="setup"/></block><l>1</l></block><script><block s="doBroadcast"><l>initialize</l><list><block s="reportJoinWords"><list><l>Layer</l><block var="i"/></list></block><block s="reportListItem"><block s="reportNewList"><list><block s="reportVariadicSum"><list><block var="i"/><l>1</l></list></block><block var="i"/></list></block><block var="setup"/></block></list></block></script></block><block s="doSetVar"><l>learning rate</l><l>0.5</l></block></script></block></script><script x="20" y="396.5000000000001"><block s="receiveGo"></block><block s="doSetVar"><l>log</l><block s="reportNewList"><list></list></block></block><block s="doRepeat"><l>2000</l><script><block s="doSetVar"><l>errors</l><l>0</l></block><block s="doForEach"><l>input</l><block s="reportListAttribute"><l><option>shuffled</option></l><block var="data"/></block><script><block s="doSetVar"><l>target</l><block s="reportListItem"><l><option>last</option></l><block var="input"/></block></block><block s="doBroadcastAndWait"><l>forward</l><list><l>Layer1</l><block var="input"/></list></block></script></block><block s="doAddToList"><block var="errors"/><block var="log"/></block><custom-block s="plot bars %l fill %n center %b"><block s="reportVariadicProduct"><list><block var="log"/><l>100</l></list></block><l>0.8</l><l><bool>false</bool></l></custom-block></script></block></script></scripts></sprite><sprite name="Layer1" idx="2" x="-66" y="-143" heading="90" scale="1" volume="100" pan="0" rotation="1" draggable="true" hidden="true" costume="0" color="155.54999999999998,9.333000000000007,0,1" pen="tip" id="364"><costumes><list struct="atomic" id="365"></list></costumes><sounds><list struct="atomic" id="366"></list></sounds><blocks></blocks><variables><variable name="weights"><list id="369"><item><list struct="atomic" id="370">0.9742746397388928,-0.7925554847898382,-0.7542952813841977</list></item><item><list struct="atomic" id="371">-0.19065307052047542,0.473460822302477,-0.20835678622724152</list></item></list></variable><variable name="inputs"><l></l></variable></variables><scripts><script x="20" y="20"><block s="receiveMessage"><l>initialize</l><list><l>shape</l></list></block><block s="doSetVar"><l>weights</l><block s="reportRandom"><l>-1.0</l><block s="reportReshape"><l>1</l><list><block s="reportListItem"><l>1</l><block var="shape"/></block><block s="reportVariadicSum"><list><block s="reportListItem"><l>2</l><block var="shape"/></block><l>1</l></list></block></list></block></block></block></script><script x="20" y="142.66666666666669"><block s="receiveMessage"><l>forward</l><list><l>sample</l></list></block><block s="doSetVar"><l>inputs</l><block var="sample"/></block><block s="doBroadcastAndWait"><l>forward</l><list><l>Layer2</l><block s="reportMap"><block s="reifyReporter"><autolambda><block s="reportMonadic"><l><option>sigmoid</option></l><block s="reportVariadicSum"><block s="reportVariadicProduct"><list><block s="reportCONS"><l>1</l><block var="inputs"/></block><l></l></list></block></block></block></autolambda><list></list></block><block var="weights"/></block></list></block></script><script x="20" y="324"><block s="receiveMessage"><l>backpropagate</l><list><l>delta</l></list></block><block s="doBroadcastAndWait"><l>backpropagate</l><list><l>ANN</l><block s="reportVariadicSum"><block s="reportVariadicProduct"><list><block s="reportMap"><block s="reifyReporter"><autolambda><block s="reportVariadicProduct"><list><block var="inputs"/><block s="reportDifference"><l>1</l><block var="inputs"/></block><l></l></list></block></autolambda><list></list></block><block var="delta"/></block><block s="reportMap"><block s="reifyReporter"><autolambda><block s="reportCDR"><l/></block></autolambda><list></list></block><block var="weights"/></block></list></block></block></list></block><block s="doChangeVar"><l>weights</l><block s="reportMap"><block s="reifyReporter"><autolambda><block s="reportVariadicProduct"><list><block s="reportCONS"><l>1</l><block var="inputs"/></block><block var="learning rate"/><l></l></list></block></autolambda><list></list></block><block var="delta"/></block></block></script></scripts></sprite><sprite name="Layer2" idx="3" x="-23" y="-144" heading="90" scale="1" volume="100" pan="0" rotation="1" draggable="true" hidden="true" costume="0" color="155.54999999999998,9.333000000000007,0,1" pen="tip" id="503"><costumes><list struct="atomic" id="504"></list></costumes><sounds><list struct="atomic" id="505"></list></sounds><blocks></blocks><variables><variable name="weights"><list id="508"><item><list struct="atomic" id="509">0.4079017419392821,0.032509940604856125,0.7988328878577118</list></item></list></variable><variable name="inputs"><l></l></variable></variables><scripts><script x="30" y="20"><block s="receiveMessage"><l>initialize</l><list><l>shape</l></list></block><block s="doSetVar"><l>weights</l><block s="reportRandom"><l>-1.0</l><block s="reportReshape"><l>1</l><list><block s="reportListItem"><l>1</l><block var="shape"/></block><block s="reportVariadicSum"><list><block s="reportListItem"><l>2</l><block var="shape"/></block><l>1</l></list></block></list></block></block></block></script><script x="30" y="142.66666666666669"><block s="receiveMessage"><l>forward</l><list><l>sample</l></list></block><block s="doSetVar"><l>inputs</l><block var="sample"/></block><block s="doBroadcastAndWait"><l>forward</l><list><l>ANN</l><block s="reportMap"><block s="reifyReporter"><autolambda><block s="reportMonadic"><l><option>sigmoid</option></l><block s="reportVariadicSum"><block s="reportVariadicProduct"><list><block s="reportCONS"><l>1</l><block var="inputs"/></block><l></l></list></block></block></block></autolambda><list></list></block><block var="weights"/></block></list></block></script><script x="30" y="324"><block s="receiveMessage"><l>backpropagate</l><list><l>delta</l></list></block><block s="doBroadcastAndWait"><l>backpropagate</l><list><l>Layer1</l><block s="reportVariadicSum"><block s="reportVariadicProduct"><list><block s="reportMap"><block s="reifyReporter"><autolambda><block s="reportVariadicProduct"><list><block var="inputs"/><block s="reportDifference"><l>1</l><block var="inputs"/></block><l></l></list></block></autolambda><list></list></block><block var="delta"/></block><block s="reportMap"><block s="reifyReporter"><autolambda><block s="reportCDR"><l/></block></autolambda><list></list></block><block var="weights"/></block></list></block></block></list></block><block s="doChangeVar"><l>weights</l><block s="reportMap"><block s="reifyReporter"><autolambda><block s="reportVariadicProduct"><list><block s="reportCONS"><l>1</l><block var="inputs"/></block><block var="learning rate"/><l></l></list></block></autolambda><list></list></block><block var="delta"/></block></block></script></scripts></sprite><watcher var="learning rate" style="normal" x="332.0000000000002" y="12.000000000000028" color="243,118,29" hidden="true"/><watcher var="bak" style="normal" x="10.000000000000455" y="10.000000000000014" color="243,118,29" hidden="true"/><watcher scope="ANN" var="log" style="normal" x="10" y="73.00000599999998" color="243,118,29" hidden="true"/><watcher scope="ANN" var="errors" style="normal" x="10" y="94.00000799999998" color="243,118,29" hidden="true"/><watcher scope="Layer2" var="delta" style="normal" x="354" y="161.00001199999997" color="243,118,29" hidden="true"/><watcher scope="ANN" var="target" style="normal" x="64" y="242.00000599999998" color="243,118,29" hidden="true"/><watcher scope="Layer2" var="inputs" style="normal" x="-3" y="219.00000999999997" color="243,118,29" hidden="true"/><watcher scope="Layer1" var="inputs" style="normal" x="80" y="64.00000399999999" color="243,118,29" hidden="true"/><watcher scope="Layer2" var="weights" style="normal" x="221" y="14.000013999999965" color="243,118,29" hidden="true"/><watcher var="weights" style="normal" x="146" y="137.00000999999997" color="243,118,29" hidden="true"/><watcher scope="Layer1" var="weights" style="normal" x="14" y="12" color="243,118,29" extX="185.5615234375" extY="52"/><watcher scope="Layer2" var="weights" style="normal" x="234" y="12.000009999999975" color="243,118,29" extX="185.5615234375" extY="35"/><watcher var="data" style="normal" x="254.00000000000023" y="102" color="243,118,29" hidden="true"/></sprites></stage><variables><variable name="data"><list id="655"><item><list struct="atomic" id="656">0,0,0</list></item><item><list struct="atomic" id="657">1,0,1</list></item><item><list struct="atomic" id="658">0,1,1</list></item><item><list struct="atomic" id="659">1,1,0</list></item></list></variable><variable name="learning rate"><l>0.5</l></variable></variables></scene></scenes></project><media name="Artificial Neural Networks - Sprite Layers But better" app="Snap! 11.0.0, https://snap.berkeley.edu" version="2"></media></snapdata>