type_agental.html

来自「一个由Mike Gashler完成的机器学习方面的includes neural」· HTML 代码 · 共 78 行

HTML
78 行
字号 
<html><head><title>Generated Documentation</title></head><body>	<image src="headerimage.png">	<br><br><table><tr><td><big><big><big style="font-family: arial;"><b>AgentAl</b></big></big></big><br>extends <a href="type_Agent.html">Agent</a><br></td><td></td></tr></table><br><br><big><big><i>Constructors (public)</i></big></big><br><div style="margin-left: 40px;"><big><b>AgentAl</b></big>(<a href="type_GAgentWorldInterface.html">GAgentWorldInterface</a>* pWorldInterface, int nShortTermMemoryCapacity, double dFocus)<br><div style="margin-left: 80px;"><font color=brown> nShortTermMemoryCapacity is the number of iterations for which the agent stores historical sense vectors dFocus specifies how many iterations the agent stays on task before it re-evaluates its goal. (1=works on the 	same goal until it achieves it, nomatter what. 0.99=stays on the same task for an average of 100 iterations. 	0.9=stays on the same task for an average of 10 iterations. etc.)</font></div><br></div><br><big><big><i>Destructors</i></big></big><br><div style="margin-left: 40px;"><big><b>~AgentAl</b></big>()<br></div><br><big><big><i>Virtual (public)</i></big></big><br><div style="margin-left: 40px;">void <big><b>Iterate</b></big>()<br></div><br><big><big><i>Protected</i></big></big><br><div style="margin-left: 40px;">void <big><b>Act</b></big>()<br><div style="margin-left: 80px;"><font color=brown> Performs the next action</font></div><br>void <big><b>AddAction</b></big>(const char* szName, IncrementalModel* ppModels)<br><div style="margin-left: 80px;"><font color=brown> ppModels should be an array of size "nSenses" (where "nSenses" is the number of senses). This method will take ownership of all the models (but not the array of pointers to them). You shouldn't add any actions until you are done adding senses.</font></div><br>void <big><b>AddSense</b></big>(const char* szName, ConstructNode* pConstruct)<br><div style="margin-left: 80px;"><font color=brown> pConstruct is the root of a construct tree for modelling this sense. This method takes ownership of pConstruct.</font></div><br>void <big><b>EvaluatePredictions</b></big>()<br><div style="margin-left: 80px;"><font color=brown> Evaluate how well action consequences were predicted</font></div><br>void <big><b>EvolveConstructs</b></big>()<br><div style="margin-left: 80px;"><font color=brown> Kill off the worst construct nodes and replace them with evolutions of the best construct nodes</font></div><br>double <big><b>GetMemoryVector</b></big>(int* i)<br><div style="margin-left: 80px;"><font color=brown> Returns a sense vector from short-term memory i = the time of the vector. 0=now, 1=just before now, 2=before that, etc.</font></div><br>void <big><b>LearnToModelTheWorld</b></big>()<br><div style="margin-left: 80px;"><font color=brown> Learn how to model the world based on sense observations</font></div><br>void <big><b>LearnToPredictHowActionsAffectTheWorld</b></big>()<br><div style="margin-left: 80px;"><font color=brown> Learn to predict how actions will affect the world</font></div><br>void <big><b>LogSpew_PrintActionConstructs</b></big>()<br>void <big><b>LogSpew_PrintConstructs</b></big>()<br>void <big><b>LogSpew_PrintObservations</b></big>()<br>void <big><b>LogSpew_PrintSenseConstructs</b></big>()<br>void <big><b>LogSpew_PrintSingleActionConstruct</b></big>(int nSense, IncrementalModel* pModel, ConstructNode* pConstruct, double* pSenseVector)<br>void <big><b>Observe</b></big>()<br><div style="margin-left: 80px;"><font color=brown> Records the sense vector into short-term memory</font></div><br>void <big><b>PredictConsequences</b></big>()<br><div style="margin-left: 80px;"><font color=brown> Predict how the world will change when the chosen action is performed</font></div><br>void <big><b>RecomputeDependencyOrder</b></big>()<br><div style="margin-left: 80px;"><font color=brown> Determines an update order for the senses that is sensitive to dependencies</font></div><br>void <big><b>SelectNextAction</b></big>()<br><div style="margin-left: 80px;"><font color=brown> Decide what to do next</font></div><br>void <big><b>SimulateAction</b></big>(int nAction, IncrementalModel* pModel, ConstructNode* pSenseConstruct)<br><div style="margin-left: 80px;"><font color=brown> Simulate the specified action on the specified sense construct</font></div><br>void <big><b>Think</b></big>()<br><div style="margin-left: 80px;"><font color=brown> Learns from observations and decides what to do next</font></div><br>void <big><b>TrainActionModel</b></big>(int nSense, double* pActionVector, double* pCurrentVector)<br><div style="margin-left: 80px;"><font color=brown> This is a helper method used by LearnToPredictHowActionsAffectTheWorld() nSense = the index of the sense pActionVector = the sense vector at the time the action was performed pCurrentVector = the sense vector at the time the consequences were observed</font></div><br></div><br></body></html>

type_agental.html - 源码说明

本页面展示了「一个由Mike Gashler完成的机器学习方面的includes neural net, naive bayesian classifier, decision tree, KNN, a genet」中的 type_agental.html 源码文件，采用 HTML 编程语言编写，共 78 行代码。您可以在线阅读完整代码内容，也可以返回资源详情页下载完整源码包进行本地学习和开发。

虫虫下载站收录了大量与classifier相关的技术资源，包括源代码、技术文档、电路图等，是电子工程师和嵌入式开发者的专业学习平台。

⌨️ 快捷键说明

复制代码Ctrl + C
搜索代码Ctrl + F
全屏模式F11
增大字号Ctrl + =
减小字号Ctrl + -
显示快捷键?