-
Notifications
You must be signed in to change notification settings - Fork 0
/
params.test.yaml
139 lines (114 loc) · 2.34 KB
/
params.test.yaml
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
settings:
n_jobs: 4
backend: joblib
mlflow_tracking_uri: http://localhost:5000
# only valid for `scripts/run_pipeline.py`
stages:
fetch_data: true
preprocess_data: true
sample_data: true
calculate_values: true
calculate_threshold_characteristics: true
evaluate_curves: true
evaluate_metrics: true
render_plots: false
time:
active: true
threshold_characteristics:
valuation_method: loo # Method used to calculate the threshold characteristics.
model: logistic_regression # Default model to use for determining the values
max_plotting_percentage: 1e-4 # Threshold for stopping plotting in direction of x-axis.
active:
experiments:
- point_removal
models:
- logistic_regression
datasets:
- diabetes
- cpu
valuation_methods:
- random
- loo
repetitions:
- 1
experiments:
point_removal:
sampler: default
curves:
accuracy_logistic_regression:
fn: metric
metric: accuracy
eval_model: logistic_regression
plots:
- accuracy
top_fraction:
fn: top_fraction
alpha_range:
from: 0.01
to: 0.5
step: 0.01
plots:
- rank_stability
value_decay:
fn: value_decay
plots:
- value_decay
metrics:
geometric_weighted_drop:
curve:
- accuracy_logistic_regression
fn: geometric_weighted_drop
input_perc: 1.0
plots:
- table
- box_wad
plots:
accuracy:
type: line
mean_agg: mean
std_agg: bootstrap
plot_perc: 0.5
x_label: "n"
y_label: "Accuracy"
table:
type: table
format: "%.3f"
box_wad:
type: boxplot
x_label: "WAD"
rank_stability:
type: line
mean_agg: intersect
x_label: "%"
y_label: "%"
value_decay:
type: line
mean_agg: mean
std_agg: bootstrap
x_label: "n"
y_label: "%"
samplers:
default:
train: 0.1667
val: 0.1667
test: 0.6667
max_samples: 3000
datasets:
diabetes:
openml_id: 37
cpu:
openml_id: 197
preprocessor:
threshold_y:
threshold: 89
models:
logistic_regression:
model: logistic_regression
solver: liblinear
valuation_methods:
random:
algorithm: random
loo:
algorithm: loo
progress: true
cache_group: acc