-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathsearch.py
157 lines (148 loc) · 6.83 KB
/
search.py
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
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
strategies = {
'FORECASTING': {
'UNIVARIATE': [
{
'strategy': 'AR',
'library': 'statsmodels'
},
*[
{
'strategy': 'SARIMAX_({},{},{})'.format(order[0], order[1], order[2]),
'library': 'statsmodels',
'order': order,
}
for order in [(1, 0, 0), (1, 1, 1), (4, 1, 2), (2, 1, 0), (0, 1, 2), (0, 1, 1), (0, 2, 2)]
],
*[
{
'strategy': 'AR_NN',
'library': 'sklearn',
'back_steps': step,
} for step in [1, 2]
],
{'strategy': 'TRA_AVERAGE', 'library': 'sklearn'},
{'strategy': 'TRA_NAIVE', 'library': 'sklearn'},
{'strategy': 'TRA_DRIFT', 'library': 'sklearn'},
],
'MULTIVARIATE': [
{
'strategy': 'VAR',
'library': 'statsmodels'
},
*[
{
'strategy': 'VAR_NN',
'library': 'sklearn',
'back_steps': step,
} for step in [1, 2, 3, 4]
],
{'strategy': 'TRA_AVERAGE', 'library': 'sklearn'},
{'strategy': 'TRA_NAIVE', 'library': 'sklearn'},
{'strategy': 'TRA_DRIFT', 'library': 'sklearn'},
]
},
'CLASSIFICATION': {
'BINARY': [
{'strategy': 'LOGISTIC_REGRESSION', 'library': 'sklearn'},
*[
{
'strategy': 'RANDOM_FOREST',
'library': 'sklearn',
'n_estimators': n_estimators
} for n_estimators in [10, 100]
],
{'strategy': 'SUPPORT_VECTOR_CLASSIFIER', 'library': 'sklearn'},
{"strategy": "RIDGE_CLASSIFIER", 'library': 'sklearn'},
{"strategy": "RIDGE_CLASSIFIER_CV", 'library': 'sklearn'},
{"strategy": "K_NEIGHBORS_CLASSIFIER", 'library': 'sklearn'},
{"strategy": "DECISION_TREE_CLASSIFIER", 'library': 'sklearn'},
{"strategy": "GRADIENT_BOOSTING_CLASSIFIER", "library": "sklearn"},
{"strategy": "LINEAR_DISCRIMINANT_ANALYSIS", "library": "sklearn"},
{"strategy": "QUADRATIC_DISCRIMINANT_ANALYSIS", "library": "sklearn"},
# {"strategy": "GAUSSIAN_PROCESS_CLASSIFIER", "library": "sklearn"},
{"strategy": "MULTINOMIAL_NAIVE_BAYES", "library": "sklearn"},
{"strategy": "GAUSSIAN_NAIVE_BAYES", "library": "sklearn"},
{"strategy": "COMPLEMENT_NAIVE_BAYES", "library": "sklearn"},
{"strategy": "ADABOOST_CLASSIFIER", "library": "sklearn"},
{'strategy': 'LOGISTIC_REGRESSION_CV', 'library': 'sklearn'},
],
'MULTICLASS': [
*[
{
'strategy': 'RANDOM_FOREST',
'library': 'sklearn',
'n_estimators': n_estimators
} for n_estimators in [10, 100]
],
{'strategy': 'SUPPORT_VECTOR_CLASSIFIER', 'library': 'sklearn'},
# {"strategy": "RIDGE_CLASSIFIER", 'library': 'sklearn'},
{"strategy": "K_NEIGHBORS_CLASSIFIER", 'library': 'sklearn'},
{"strategy": "DECISION_TREE_CLASSIFIER", 'library': 'sklearn'},
{"strategy": "GRADIENT_BOOSTING_CLASSIFIER", "library": "sklearn"},
{"strategy": "LINEAR_DISCRIMINANT_ANALYSIS", "library": "sklearn"},
{"strategy": "QUADRATIC_DISCRIMINANT_ANALYSIS", "library": "sklearn"},
# {"strategy": "GAUSSIAN_PROCESS_CLASSIFIER", "library": "sklearn"},
{"strategy": "MULTINOMIAL_NAIVE_BAYES", "library": "sklearn"},
{"strategy": "GAUSSIAN_NAIVE_BAYES", "library": "sklearn"},
{"strategy": "COMPLEMENT_NAIVE_BAYES", "library": "sklearn"},
{"strategy": "ADABOOST_CLASSIFIER", "library": "sklearn"},
],
'MULTILABEL': [
*[
{
'strategy': 'RANDOM_FOREST',
'library': 'sklearn',
'n_estimators': n_estimators
} for n_estimators in [10, 100]
],
{'strategy': 'SUPPORT_VECTOR_CLASSIFIER', 'library': 'sklearn'},
{"strategy": "RIDGE_CLASSIFIER", 'library': 'sklearn'},
{"strategy": "K_NEIGHBORS_CLASSIFIER", 'library': 'sklearn'},
{"strategy": "DECISION_TREE_CLASSIFIER", 'library': 'sklearn'},
]
},
'REGRESSION': {
'UNIVARIATE': [
{'strategy': 'ORDINARY_LEAST_SQUARES', 'library': 'sklearn'},
{"strategy": "RANDOM_FOREST_REGRESSOR", 'library': 'sklearn'},
{"strategy": "SUPPORT_VECTOR_REGRESSION", 'library': 'sklearn'},
{"strategy": "K_NEIGHBORS_REGRESSOR", 'library': 'sklearn'},
{"strategy": "DECISION_TREE_REGRESSOR", 'library': 'sklearn'},
{"strategy": "LASSO_REGRESSION", "library": "sklearn"},
{"strategy": "LASSO_REGRESSION_LARS", "library": "sklearn"},
{"strategy": "ELASTIC_NET", "library": "sklearn"},
{"strategy": "ORTHOGONAL_MATCHING", "library": "sklearn"},
{"strategy": "ADABOOST_REGRESSOR", "library": "sklearn"},
{"strategy": "GRADIENT_BOOSTING_REGRESSOR", "library": "sklearn"},
# {"strategy": "GAUSSIAN_PROCESS_REGRESSOR", "library": "sklearn"},
{"strategy": "RIDGE_CV", "library": "sklearn"},
]
}
}
class SearchManager(object):
def __init__(self, system_params, problem_specification):
self.problem_specification = problem_specification
self.system_params = system_params
task = problem_specification['taskType']
subtask = problem_specification.get('taskSubtype')
# TODO: forecasting subtypes need rework
if problem_specification['taskType'] == 'FORECASTING':
variables = problem_specification['targets'] + problem_specification['predictors']
tmp_cross = problem_specification.get('crossSection', [])
variables = [var for var in variables if var not in tmp_cross]
subtask = 'MULTIVARIATE' if len(variables) > 2 else 'UNIVARIATE'
# if problem_specification['taskType'] == 'FORECASTING' and self.problem_specification.get('crossSection'):
self.generator = iter(strategies.get(task, {}).get(subtask, []))
# else:
# self.generator = iter(strategies.get(task, {}).get(subtask, []))
def get_pipeline_specification(self):
try:
model_specification = next(self.generator)
except StopIteration:
return
return {
'preprocess': None,
'model': model_specification
}
def metalearn_result(self, pipeline_specification, scores):
pass