Changeset 11710 for branches/HeuristicLab.Problems.GrammaticalOptimization/HeuristicLab.Algorithms.Bandits
- Timestamp:
- 12/21/14 09:19:54 (9 years ago)
- Location:
- branches/HeuristicLab.Problems.GrammaticalOptimization/HeuristicLab.Algorithms.Bandits
- Files:
-
- 7 added
- 1 deleted
- 2 edited
Legend:
- Unmodified
- Added
- Removed
-
branches/HeuristicLab.Problems.GrammaticalOptimization/HeuristicLab.Algorithms.Bandits/EpsGreedyPolicy.cs
r11708 r11710 24 24 var maxReward = double.NegativeInfinity; 25 25 int bestAction = -1; 26 int curAction = -1;27 foreach (var avgReward in sumReward.Zip(tries, (r, t) => r / (t + 1))) { // prevent division by zero28 curAction++;26 for (int i = 0; i < NumActions; i++) { 27 if (tries[i] == 0) return i; 28 var avgReward = sumReward[i] / tries[i]; 29 29 if (maxReward < avgReward) { 30 30 maxReward = avgReward; 31 bestAction = curAction;31 bestAction = i; 32 32 } 33 33 } -
branches/HeuristicLab.Problems.GrammaticalOptimization/HeuristicLab.Algorithms.Bandits/HeuristicLab.Algorithms.Bandits.csproj
r11708 r11710 40 40 </ItemGroup> 41 41 <ItemGroup> 42 <Compile Include="B andit.cs" />42 <Compile Include="BernoulliBandit.cs" /> 43 43 <Compile Include="BanditPolicy.cs" /> 44 <Compile Include="TruncatedNormalBandit.cs" /> 45 <Compile Include="UCBNormalPolicy.cs" /> 46 <Compile Include="UCB1TunedPolicy.cs" /> 47 <Compile Include="UCB1Policy.cs" /> 44 48 <Compile Include="EpsGreedyPolicy.cs" /> 45 49 <Compile Include="IPolicy.cs" /> 46 50 <Compile Include="Properties\AssemblyInfo.cs" /> 47 51 <Compile Include="RandomPolicy.cs" /> 52 </ItemGroup> 53 <ItemGroup> 54 <Folder Include="Bandits\" /> 55 <Folder Include="Policies\" /> 48 56 </ItemGroup> 49 57 <Import Project="$(MSBuildToolsPath)\Microsoft.CSharp.targets" />
Note: See TracChangeset
for help on using the changeset viewer.