Last change
on this file since 13475 was
12893,
checked in by gkronber, 9 years ago
|
#2283: experiments on grammatical optimization algorithms (maxreward instead of avg reward, ...)
|
File size:
1.1 KB
|
Line | |
---|
1 | using System;
|
---|
2 | using System.Collections.Generic;
|
---|
3 | using System.Diagnostics;
|
---|
4 | using System.Linq;
|
---|
5 | using System.Text;
|
---|
6 | using System.Threading.Tasks;
|
---|
7 | using HeuristicLab.Common;
|
---|
8 |
|
---|
9 | namespace HeuristicLab.Algorithms.Bandits.BanditPolicies {
|
---|
10 | public class BernoulliPolicyActionInfo : IBanditPolicyActionInfo {
|
---|
11 | public int NumSuccess { get; private set; }
|
---|
12 | public int NumFailure { get; private set; }
|
---|
13 | public int Tries { get { return NumSuccess + NumFailure; } }
|
---|
14 | public double MaxReward { get; private set; }
|
---|
15 | public double Value {
|
---|
16 | get {
|
---|
17 | return NumSuccess / (double)(Tries);
|
---|
18 | }
|
---|
19 | }
|
---|
20 | public void UpdateReward(double reward) {
|
---|
21 | //Debug.Assert(reward.IsAlmost(0.0) || reward.IsAlmost(1.0));
|
---|
22 |
|
---|
23 | //if (reward.IsAlmost(1.0)) NumSuccess++;
|
---|
24 | MaxReward = Math.Max(MaxReward, reward);
|
---|
25 | if (reward > 0) NumSuccess++;
|
---|
26 | else NumFailure++;
|
---|
27 | }
|
---|
28 | public void Reset() {
|
---|
29 | NumSuccess = 0;
|
---|
30 | NumFailure = 0;
|
---|
31 | MaxReward = double.NegativeInfinity;
|
---|
32 |
|
---|
33 | }
|
---|
34 | public void PrintStats() {
|
---|
35 | Console.WriteLine("expected value {0,5:F2}", Value);
|
---|
36 | }
|
---|
37 | }
|
---|
38 | }
|
---|
Note: See
TracBrowser
for help on using the repository browser.