Recursive Bellman Variance Targets / results.json
Failed on benchmark
1{
2 "seed": 924,
3 "parameters": {
4 "gamma": 0.9,
5 "reward_sigma": 0.35
6 },
7 "prediction_1_recursive_vs_mc": [
8 {
9 "h": 1,
10 "recursive_var": 0.12249999999999997,
11 "mc_var": 0.12318373640244373,
12 "relative_error": 0.005581521652602149
13 },
14 {
15 "h": 2,
16 "recursive_var": 0.9588249999999998,
17 "mc_var": 0.9588483281225746,
18 "relative_error": 2.4329906473832865e-05
19 },
20 {
21 "h": 4,
22 "recursive_var": 3.143065340611,
23 "mc_var": 3.1388989964166893,
24 "relative_error": 0.0013255671590653535
25 },
26 {
27 "h": 8,
28 "recursive_var": 5.854791020556265,
29 "mc_var": 5.839320859524046,
30 "relative_error": 0.002642307979550974
31 },
32 {
33 "h": 12,
34 "recursive_var": 7.033590576656854,
35 "mc_var": 7.046224033481828,
36 "relative_error": 0.0017961603945076865
37 }
38 ],
39 "prediction_2_selection_term": [
40 {
41 "p": 0.0,
42 "predicted": 0.0,
43 "measured": 0.0
44 },
45 {
46 "p": 0.1,
47 "predicted": 2.2500000000000004,
48 "measured": 2.25
49 },
50 {
51 "p": 0.25,
52 "predicted": 4.6875,
53 "measured": 4.6875
54 },
55 {
56 "p": 0.5,
57 "predicted": 6.25,
58 "measured": 6.25
59 },
60 {
61 "p": 0.75,
62 "predicted": 4.6875,
63 "measured": 4.6875
64 },
65 {
66 "p": 0.9,
67 "predicted": 2.2499999999999996,
68 "measured": 2.2499999999999996
69 },
70 {
71 "p": 1.0,
72 "predicted": 0.0,
73 "measured": 0.0
74 }
75 ],
76 "prediction_3_horizon": [
77 {
78 "transition_randomness": 0.0,
79 "points": [
80 {
81 "h": 1,
82 "predicted": 0.12249999999999998,
83 "mc_var": 0.12256535678062151
84 },
85 {
86 "h": 2,
87 "predicted": 0.22172499999999998,
88 "mc_var": 0.2214153524203679
89 },
90 {
91 "h": 4,
92 "predicted": 0.3671987725,
93 "mc_var": 0.3667119714213641
94 },
95 {
96 "h": 8,
97 "predicted": 0.5252658036134997,
98 "mc_var": 0.5271379751040887
99 },
100 {
101 "h": 16,
102 "predicted": 0.6225986174744299,
103 "mc_var": 0.6250887069801233
104 }
105 ]
106 },
107 {
108 "transition_randomness": 1.0,
109 "points": [
110 {
111 "h": 1,
112 "predicted": 1.1225,
113 "mc_var": 1.1223659783526903
114 },
115 {
116 "h": 2,
117 "predicted": 2.0317250000000002,
118 "mc_var": 2.027325707781389
119 },
120 {
121 "h": 4,
122 "predicted": 3.3647397725,
123 "mc_var": 3.3712183120062322
124 },
125 {
126 "h": 8,
127 "predicted": 4.813149914744111,
128 "mc_var": 4.812612597840098
129 },
130 {
131 "h": 16,
132 "predicted": 5.705036311143246,
133 "mc_var": 5.707145087263986
134 }
135 ]
136 }
137 ],
138 "mini_experiment": {
139 "device": "cuda",
140 "baseline_mse": 0.2465386539697647,
141 "baseline_calibration_ratio": 0.2702868580818176,
142 "idea_mse": 0.24735412001609802,
143 "idea_calibration_ratio": 0.2707649767398834
144 }
145}