From 98139f573eb179c8f5a06ba6c8d8883376814ccf Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?=C3=89tienne=20Simon?= Date: Thu, 2 Jul 2015 12:59:15 -0400 Subject: Remove _simple --- config/joint_mlp_tgtcls_1_cswdtx_bigger.py | 54 ++++++++++++++++++++++++++++++ 1 file changed, 54 insertions(+) create mode 100644 config/joint_mlp_tgtcls_1_cswdtx_bigger.py (limited to 'config/joint_mlp_tgtcls_1_cswdtx_bigger.py') diff --git a/config/joint_mlp_tgtcls_1_cswdtx_bigger.py b/config/joint_mlp_tgtcls_1_cswdtx_bigger.py new file mode 100644 index 0000000..72d3c83 --- /dev/null +++ b/config/joint_mlp_tgtcls_1_cswdtx_bigger.py @@ -0,0 +1,54 @@ +import os +import cPickle + +from blocks.initialization import IsotropicGaussian, Constant + +import data +from model.joint_mlp_tgtcls import Model, Stream + + +n_begin_end_pts = 7 # how many points we consider at the beginning and end of the known trajectory + +with open(os.path.join(data.path, 'arrival-clusters.pkl')) as f: + dest_tgtcls = cPickle.load(f) + +# generate target classes for time prediction as a Fibonacci sequence +time_tgtcls = [1, 2] +for i in range(21): + time_tgtcls.append(time_tgtcls[-1] + time_tgtcls[-2]) + +dim_embeddings = [ + ('origin_call', data.origin_call_size, 15), + ('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), +] + +# Common network part +dim_input = n_begin_end_pts * 2 * 2 + sum(x for (_, _, x) in dim_embeddings) +dim_hidden = [5000] + +# Destination prediction part +dim_hidden_dest = [] +dim_output_dest = dest_tgtcls.shape[0] + +# Time prediction part +dim_hidden_time = [] +dim_output_time = len(time_tgtcls) + +# Cost ratio between distance cost and time cost +time_cost_factor = 4 + +embed_weights_init = IsotropicGaussian(0.01) +mlp_weights_init = IsotropicGaussian(0.1) +mlp_biases_init = Constant(0.01) + +learning_rate = 0.0001 +momentum = 0.99 +batch_size = 200 + +valid_set = 'cuts/test_times_0' +max_splits = 100 -- cgit v1.2.3