xanderdunn/reinforcementLearning

Issues

• Add Benchmarks

  #35 opened 9 years ago
  0

• Implement ScalaDoc

  #33 opened 9 years ago
  0

• Add WartRemover static analyzer

  #32 opened 9 years ago
  0

• Fix all Codacy Issues

  #31 opened 9 years ago
  0

• Add Scapegoat Scala Static Analyzer

  #30 opened 9 years ago
  0

• Use the neural net with 9 outputs where each spits out the value of taking that action in the given state

  #29 opened 9 years ago
  0

• Add an exception to test that the agents aren't playing the same game over and over

  #28 opened 9 years ago
  0

• Choose more intelligent hidden and output activation functions

  #27 opened 9 years ago
  0

• Add neural net unit tests across larger and negative domains and ranges

  #26 opened 9 years ago
  0

• Add code coverage statistics

  #25 opened 9 years ago
  0

• Use the cross entropy cost function

  #24 opened 9 years ago
  1

• Turn the README into a readable paper on methods and results with in-line graphs

  #23 opened 9 years ago
  0

• Use the cross entropy cost function

  #22 opened 9 years ago
  0

• Add batch training

  #21 opened 9 years ago
  0

• Convert the gridworld project over to scalatest

  #20 opened 9 years ago
  0

• Use non-constant alpha

  #19 opened 9 years ago
  0

• Fix the use of the bipolar sigmoid function

  #18 opened 9 years ago
  0

• Try Quickprop or RPROP instead of Backprop

  #17 opened 9 years ago
  0

• Simplify all the if()

  #16 opened 9 years ago
  0

• Add linear function approximators for the value function

  #15 opened 9 years ago
  0

• Fix graph output and switch over to an external plotting library

  #14 opened 9 years ago
  0

• Make the Parameters part of GameParameters

  #13 opened 9 years ago
  0

• Compare Tabular and Neural Net Q(s, a)

  #12 opened 9 years ago
  0

• Use Java standard exceptions instead of your own custom exceptions

  #11 opened 9 years ago
  0

• Test the convergence of Q values

  #10 opened 9 years ago
  0

• Add pursuit methods

  #9 opened 9 years ago
  0

• Add reinforcement comparison

  #8 opened 9 years ago
  0

• Try alpha = 1/k

  #7 opened 9 years ago
  0

• Add the Q Learning variation on page 151

  #6 opened 9 years ago
  0

• Add SARSA(lambda)

  #5 opened 9 years ago
  0

• Add option to use a single neural network for both Agents

  #4 opened 9 years ago
  0

• Use a single neural network for each Agent

  #3 opened 9 years ago
  0

• Turn off learning altogether

  #2 opened 9 years ago
  0

• Add exception on number of times rewarded per episode

  #1 opened 9 years ago
  0