def train_oneslice_model(areas_all,train_true):
#train OneSliceModel using CNN_B result
print(" --- train oneslice model :");
oneslice_model = OneSliceModel();
oneslice_model.fit(areas_all,train_true);
oneslice_predict = oneslice_model.predict(areas_all);
analysis.evaluate_pred(oneslice_predict, train_true);
return oneslice_predict;
def train_sex_age_model(info, train_true):
##train sex_age model
print(" ------ train sex age model :");
sa_model = SexAgeModel();
sa_model.fit(info,train_true);
sa_predict = sa_model.predict(info);
analysis.evaluate_pred(sa_predict, train_true);
return sa_predict;
def train_ch4_model(ch4_data, train_true):
#tencia's 4-ch model, implemented 02/27
print(" ---- train ch4 model :");
ch4_model = Ch4Model();
ch4_model.fit(ch4_data, train_true);
ch4_pred = ch4_model.predict(ch4_data);
analysis.evaluate_pred(ch4_pred, train_true);
return ch4_pred;
def train_sax_cnt_filter_model(areas_all, cont_all, train_true,cleaner=[]):
print(" ---- train sax countour filter model :");
cnt_result = analysis.get_preliminary_volume_cnt_filter(areas_all,cont_all,cleaner=cleaner);
cnt_sax_model = SaxFilterModel();
cnt_sax_model.fit(cnt_result,train_true);
cnt_sax_predict = cnt_sax_model.predict(cnt_result);
analysis.evaluate_pred(cnt_sax_predict, train_true);
return cnt_sax_predict;
def train_sax_model(areas_all,train_true, version,cleaner=[]):
print(" ---- train sax model :");
sax_model = SaxModel(version=version);
result = analysis.get_preliminary_volume(areas_all,cleaner=cleaner);
sax_model.fit(result,train_true);
sax_predict = sax_model.predict(result);
analysis.evaluate_pred(sax_predict, train_true);
return sax_predict;
def train_sax_cnt_model(areas_all, cont_all, train_true,version=2,cleaner=[]):
#sax model method2, use contour compeleteness to filter result
print(" ---- train sax countour model :");
cnt_sax_model = SaxModel(version=version);
cnt_result = analysis.get_preliminary_volume_cnt(areas_all, cont_all,cleaner=cleaner);
cnt_sax_model.fit(cnt_result,train_true);
cnt_sax_predict = cnt_sax_model.predict(cnt_result);
analysis.evaluate_pred(cnt_sax_predict, train_true);
return cnt_sax_predict;
def train_oneslice_model(areas_all, train_true):
#train OneSliceModel using CNN_B result
turn = dhl.turn()
dhl.log_time_stamp(turn, 'train_pred.py', '-', 'train_oneslice_model',
dhl.START_FUNCTION, ' - ')
oneslice_model = OneSliceModel()
oneslice_model.fit(areas_all, train_true)
oneslice_predict = oneslice_model.predict(areas_all)
analysis.evaluate_pred(oneslice_predict, train_true)
dhl.log_time_stamp(turn, 'train_pred.py', '-', 'train_oneslice_model',
dhl.END_FUNCTION, ' - ')
return oneslice_predict
def train_sex_age_model(info, train_true):
##train sex_age model
turn = dhl.turn()
dhl.log_time_stamp(turn, 'train_pred.py', '-', 'train_sex_age_model',
dhl.START_FUNCTION, ' - ')
sa_model = SexAgeModel()
sa_model.fit(info, train_true)
sa_predict = sa_model.predict(info)
analysis.evaluate_pred(sa_predict, train_true)
dhl.log_time_stamp(turn, 'train_pred.py', '-', 'train_sex_age_model',
dhl.END_FUNCTION, ' - ')
return sa_predict
def train_sax_model(areas_all, train_true, version, cleaner=[]):
turn = dhl.turn()
dhl.log_time_stamp(turn, 'train_pred.py', '-', 'train_sax_model',
dhl.START_FUNCTION, ' - ')
sax_model = SaxModel(version=version)
result = analysis.get_preliminary_volume(areas_all, cleaner=cleaner)
sax_model.fit(result, train_true)
sax_predict = sax_model.predict(result)
analysis.evaluate_pred(sax_predict, train_true)
dhl.log_time_stamp(turn, 'train_pred.py', '-', 'train_sax_model',
dhl.END_FUNCTION, ' - ')
return sax_predict
def train_ch4_model(ch4_data, train_true):
#tencia's 4-ch model, implemented 02/27
turn = dhl.turn()
dhl.log_time_stamp(turn, 'train_pred.py', '-', 'train_ch4_model',
dhl.START_FUNCTION, ' - ')
ch4_model = Ch4Model()
ch4_model.fit(ch4_data, train_true)
ch4_pred = ch4_model.predict(ch4_data)
analysis.evaluate_pred(ch4_pred, train_true)
dhl.log_time_stamp(turn, 'train_pred.py', '-', 'train_ch4_model',
dhl.END_FUNCTION, ' - ')
return ch4_pred
def train_sax_cnt_filter_model(areas_all, cont_all, train_true, cleaner=[]):
turn = dhl.turn()
dhl.log_time_stamp(turn, 'train_pred.py', '-',
'train_sax_cnt_filter_model', dhl.START_FUNCTION, ' - ')
cnt_result = analysis.get_preliminary_volume_cnt_filter(areas_all,
cont_all,
cleaner=cleaner)
cnt_sax_model = SaxFilterModel()
cnt_sax_model.fit(cnt_result, train_true)
cnt_sax_predict = cnt_sax_model.predict(cnt_result)
analysis.evaluate_pred(cnt_sax_predict, train_true)
dhl.log_time_stamp(turn, 'train_pred.py', '-',
'train_sax_cnt_filter_model', dhl.END_FUNCTION, ' - ')
return cnt_sax_predict
def train_sax_cnt_model(areas_all,
cont_all,
train_true,
version=2,
cleaner=[]):
#sax model method2, use contour compeleteness to filter result
turn = dhl.turn()
dhl.log_time_stamp(turn, 'train_pred.py', '-', 'train_sax_cnt_model',
dhl.START_FUNCTION, ' - ')
cnt_sax_model = SaxModel(version=version)
cnt_result = analysis.get_preliminary_volume_cnt(areas_all,
cont_all,
cleaner=cleaner)
cnt_sax_model.fit(cnt_result, train_true)
cnt_sax_predict = cnt_sax_model.predict(cnt_result)
analysis.evaluate_pred(cnt_sax_predict, train_true)
dhl.log_time_stamp(turn, 'train_pred.py', '-', 'train_sax_cnt_model',
dhl.END_FUNCTION, ' - ')
return cnt_sax_predict
Ntrain = train_true.shape[0];#500, becomes 700 when validate data released
print("number of train cases is {}".format(Ntrain));
filter_ll = -2000;
#### train models,
########### default models
sa_predict = train_sex_age_model(info, train_true);
ch4_predict = train_ch4_model(ch4_data, train_true);
pick = [0,1];
qi_best,qi_best_cont = analysis.take_best_contour([qi_areas[i] for i in pick],[qi_cnts[i] for i in pick],method=1, filter_ll = filter_ll);
oneslice_pred = train_oneslice_model(qi_best,train_true);
# fit the fall back model, a combination of oneslice_model and 4-ch model,
# if it still fails use the sex-age model
# 0.6 * oneslice_predict + 0.4 * ch4_predict; (use fixed 0.6, 0.4)
default_pred = build_default_model(oneslice_pred, ch4_predict, sa_predict);
analysis.evaluate_pred(default_pred, train_true);
########## sax based CNN models
pick = [0,1];
qi_best,qi_best_cont = analysis.take_best_contour([qi_areas[i] for i in pick],[qi_cnts[i] for i in pick],method=1, filter_ll=filter_ll);
qi_sax_pred = train_sax_model(qi_best,train_true, version = 2, cleaner = cleaner);
qi_sax_cnt_pred = train_sax_cnt_model(qi_best, qi_best_cont, train_true,version=2, cleaner = cleaner);
qi_sax_filter_pred = train_sax_cnt_filter_model(qi_best,qi_best_cont,train_true, cleaner = cleaner);
# fit the combined model based on the cnn-sax models.
# when it fails, fall to the previously fitted fall back model
print(" --------- average models --");
#Submit V9
all_models = [qi_sax_pred, qi_sax_cnt_pred, qi_sax_filter_pred];
ave_model = AverageModel(ll = 1.0e-4);