Man pages for MartinKies/RLR
Reinforcement Learning with R

Act.A3CDetermines which action the algorithm takes
Action.2.Choice.PDAction to Array for Prisoners Dilemma
Action.Encoding.Info.PDGet Info of Action Encoding
Action.Encoding.Info.Simple.GameGet Info of Action Encoding
Act.QLearningPersExpPathDetermines which action to take
Advantage.functionCalculates N-Step Returns or weighted Temporal Difference...
Choice.2.Action.PDArray to Action for Prisoners Dilemma
Choice.2.Action.Simple.GameArray to Action for Simple Game
compare.explorationStrategy highlighting differences in Exploration states are...
Convert.2.trainConverts stored Memory into arrays.
counter.strat.aA best answer strategy
counter.strat.cA bast answer to a testpool strategy
counter.strat.dA best answer strategy
counter.strat.eA best answer strategy
counter.strat.fA best answer strategy
counter.strat.gA best answer strategy
counter.strat.hA best answer strategy
counter.strat.iA best answer strat
Define_GraphGraph for Network Loss according to A3C.
Define_Graph_Gradient_UpdateGraph to update Network weights
Discounted.Reward.PDUpdate Score based on expected Value of reward
Encode.Game.States.PDTransforms List of Gamestates to std encoding form
Encoding.Harper.PDEncoding based on Reinforcement Learning Produces Dominant...
Encoding.last.X.rounds.PDFlexible Encoding Function which expects the following...
Encoding.Manager.PDInternal Function to make working with different encodings...
Extend.Memory.QLearningPersExpPathExtend Memory by specified experiences
External.Eval.PDEvaluate the current strategy using StratTourn
fix.price.locExample srategy for the Hotelling game
Generate.Start.State.PDGenerates Start State for Prisoners Dilemma Game
Generate.Start.State.Simple.GameGenerates Start State for Simple Game
get.against.itself.benchmarkPayoff of strategy against itself
get.antistratGet vector of names of counter strategy
get.benchmarkPayoff of best answer against the strategy
get.conversionConversion factor
Get.Def.Par.A3CGet Default Parameters of A3C.
Get.Def.Par.Neural.NetworkDefine default Parameters of the Neural Network Function
Get.Def.Par.Neural.Network.A3CGet Default Parameters of the Feed-Forward Neural Network for...
Get.Def.Par.Neural.Network.A3C.LSTMGet Default Parameters of the LSTM Neural Network for the A3C...
Get.Def.Par.NN.Legacy.Thesis.BasicDefault Parameters Basic Neural Network as used for the...
Get.Def.Par.NN.Legacy.v.0.1.6Default Parameters Neural Network v.0.1.6
Get.Def.Par.QLearningPersExpPathDefault Parameters for (improved) Q-Learning
Get.Def.Par.QLearningPersExpPath.Legacy.ThesisOpt.RNNDefault Parameters QLearningPersExpPath Recurrent NN v.0.1.6
Get.Def.Par.QLearningPersExpPath.Legacy.ThesisOpt.XGBDefault Parameters QLearningPersExpPath of the thesis of...
Get.Def.Par.QLearningPersExpPath.Legacy.v.0.1.6Default Parameters QLearningPersExpPath v.0.1.6
Get.Def.Par.QLearningPersExpPath.QLearning.BasicDefault Parameters normal Q-Learnung used in the thesis of...
Get.Def.Par.RNNDefine default Parameters of the RNN Function
Get.Def.Par.RNN.Legacy.ThesisOptDefault Parameters Recurrent Neural Network Thesis Martin...
Get.Def.Par.RNN.Legacy.v.0.1.6Default Parameters Recurrent Neural Network version 0.1.6
Get.Def.Par.XGBoostDefault Parameters for XGBoost
Get.Def.Par.XGBoost.Legacy.ThesisOptDefault Parameters Gradient Boosting Thesis Martin Kies
Get.Def.Par.XGBoost.Legacy.v.0.1.6Default Parameters Gradient Boosting version 0.1.6
Get.Game.Object.PDGet Game Object which fully defines Prisoners Dilemma.
Get.Game.Object.Simple.GameGet Game Object which fully defines simple game.
Get.Game.Param.PDDefines the game environment of the IPD
Get.Game.Param.PD.Legacy.BattleOfStrategies2013.BaselineDefault Parameters Game - Rules of Seminar 2013
Get.Game.Param.PD.Legacy.BattleOfStrategies2019Default Parameters Game - Rules of Seminar 2019
Get.Par.PDDefines model parameters for 'Prisoners Dilemma'
Get.Par.Simple.GameDefines model parameters for 'Simple Game'
Initialise.A3CSet changeable A3C Parameters.
Initialise.QLearningPersExpPathSet changeable model variables
Memory.Random.Play.PDGenerate Memory where strategies play against a random...
Memory.Self.Play.PDGenerate Memory where strategies play against themselves
Model.strat.maximum.full.TenA strategy to be used after model has been trained
net.nice0A prof strategy
net.nice1A variant to net.nice0
net.nice.minus1A variant to net.nice0
net.nice.start1A variant to net.nice0
NN.strat.full.zeroA strategy to be used after model has been trained
NN.strat.mainThe actual strategy after model has been trained
NN.strat.Slim.TenTenA strategy to be used after model has been trained
NN.strat.Slim.TenTen.QLearningA strategy to be used after model has been trained
NN.strat.static.end.TenA strategy to be used after model has been trained
PID.controllerPID controller, which generates ouptput based on the error
Play.Multiple.Games.QLearningPersExpPathTrain multiple games
Play.On.Strategy.QLearningPersExpPathPlay the game based on strategy
Predict.Neural.NetworkEvaluate Neural Network
Predict.Neural.Network.A3CPredict Neural Network
Predict.RNNEvaluate Recurrent RNN Worker Memory to visualize with shiny
Q.on.hist.PD.QLearningQ-values based on history of IPD
redim.stateChange dimensionality of the state array.
Replay.QLearningPersExpPathTrain model of Q learning
Setup.Neural.NetworkSetup a Neural Network
Setup.Neural.Network.A3CSetup a Feed-Forward Neural Network for the...
Setup.Neural.Network.A3C.LSTMSetup a Neural Network with an LSTM-Layer for the...
Setup.QLearningPersExpPathSets up a model based on model parameters
Setup.RNNSetup a RNN
smooth.averageCalculates a sensible moving average based on smoothing...
smooth.triangleCalculates a sensible moving average based on smoothing...
State.2.Array.PDState to Array for Prisoners Dilemma
State.2.Array.Simple.GameState to Array for Simple Game
State.Transition.PDGet next State of Prisoners Dilemma Game
State.Transition.Simple.GameGet next State of Simple Game
strat.aA testpool strategy
strat.bA testpool strategy
strat.cA testpool strategy
strat.dA testpool strategy
strat.eA testpool strategy
strat.fA testpool strategy
strat.gA testpool strategy
strat.hA testpool strategy
strat.iA testpool strategy
Train.A3cUse the A3C algorithm to train a model
Train.Neural.NetworkTrain Neural Network
Train.On.Memory.QLearningPersExpPathTrains model based on memory
Train.QLearningPersExpPathTrain a model based on Q-Learning
Train.RNNTrain RNN
Update.Evaluator.QLearningPersExpPathControlled Copying of Models
Update.Memory.QLearningPersExpPathAdd historic Q-Values and Curiosity to memory
Weighted.DiscountCalculates a weighted Mean
Worker.A3CDefines an Agent based on the A3C-Algorithm
MartinKies/RLR documentation built on Dec. 24, 2019, 10:02 p.m.