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
|
import os
import cPickle
from blocks.initialization import IsotropicGaussian, Constant
import data
from model.dest_simple_mlp_tgtcls import Model, Stream
n_begin_end_pts = 5 # how many points we consider at the beginning and end of the known trajectory
n_valid = 1000
with open(os.path.join(data.path, 'arrival-clusters.pkl')) as f: tgtcls = cPickle.load(f)
dim_embeddings = [
('origin_call', data.origin_call_train_size, 10),
('origin_stand', data.stands_size, 10),
('week_of_year', 52, 10),
('day_of_week', 7, 10),
('qhour_of_day', 24 * 4, 10),
('day_type', 3, 10),
('taxi_id', 448, 10),
]
dim_input = n_begin_end_pts * 2 * 2 + sum(x for (_, _, x) in dim_embeddings)
dim_hidden = [500]
dim_output = tgtcls.shape[0]
embed_weights_init = IsotropicGaussian(0.001)
mlp_weights_init = IsotropicGaussian(0.01)
mlp_biases_init = Constant(0.001)
learning_rate = 0.0001
momentum = 0.99
batch_size = 100
use_cuts_for_training = True
max_splits = 1
valid_set = 'cuts/test_times_0'
|
