In [12]:
top_n = plot_feature_importance(model.feature_importances_, features,'CatBoost', n=80, figsize=(20,12))
print(f'Mean CV score for {MAX_DEPTH} max_depth: {np.mean(score_list)}')
Mean CV score for 10 max_depth: 0.7933820749798006
import random
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
import scipy.stats
from sklearn.model_selection import KFold, StratifiedKFold, train_test_split
from sklearn.metrics import roc_auc_score
from catboost import CatBoostClassifier
import gc
import datetime
from IPython.display import display
import warnings
import vaex
vaex.multithreading.thread_count_default = 8
import vaex.ml
from cycler import cycler
from colorama import Fore, Back, Style
# config
MAX_DEPTH = 10
ITERATIONS = 28000
DATA_PATH = '../input/amex-data-integer-dtypes-parquet-format/' # denoised data from raddar
LABELS_PATH = '../input/amex-default-prediction/train_labels.csv' # original data sources
TEST_FEAT_PATH = '../input/amex-denoised-aggregated-features/test_feat.parquet' # aggregated features I shared
TRAIN_FEAT_PATH = '../input/amex-denoised-aggregated-features/train_feat.parquet'
def plot_feature_importance(importance,names,model_type, n=50, figsize=(16,10)):
# 특징명과 특징의 중요도로 배열을 만든다.
feature_importance = np.array(importance)
feature_names = np.array(names)
# 딕셔너리를 이용하여 데이터 프레임을 생성
data={'feature_names':feature_names,'feature_importance':feature_importance}
fi_df = pd.DataFrame(data)
# 특징 중요도를 정렬
fi_df.sort_values(by=['feature_importance'], ascending=False,inplace=True)
fi_df = fi_df[:n]
# Define size of bar plot
plt.figure(figsize=figsize)
# Plot Searborn bar chart
sns.barplot(x=fi_df['feature_importance'], y=fi_df['feature_names'])
# 그래프의 레이브를 추가
plt.title(model_type + ' FEATURE IMPORTANCE')
plt.xlabel('FEATURE IMPORTANCE')
plt.ylabel('FEATURE NAMES')
plt.tight_layout()
plt.show()
return fi_df.feature_names
def amex_metric(y_true: np.array, y_pred: np.array) -> float:
# count of positives and negatives
n_pos = y_true.sum()
n_neg = y_true.shape[0] - n_pos
# sorting by describing prediction values
indices = np.argsort(y_pred)[::-1]
preds, target = y_pred[indices], y_true[indices]
# filter the top 4% by cumulative row weights
weight = 20.0 - target * 19.0
cum_norm_weight = (weight / weight.sum()).cumsum()
four_pct_filter = cum_norm_weight <= 0.04
# default rate captured at 4%
d = target[four_pct_filter].sum() / n_pos
# weighted Gini coefficient
lorentz = (target / n_pos).cumsum()
gini = ((lorentz - cum_norm_weight) * weight).sum()
# max weighted Gini coefficient
gini_max = 10 * n_neg * (1 - 19 / (n_pos + 20 * n_neg))
# normalized weighted Gini coefficient
g = gini / gini_max
return 0.5 * (g + d)
# TEST_FEAT_PATH = '../input/amex-denoised-aggregated-features/test_feat.parquet'
# TRAIN_FEAT_PATH = '../input/amex-denoised-aggregated-features/train_feat.parquet'
def get_data(read_from_cache=True):
train = pd.read_parquet(TRAIN_FEAT_PATH)
test = pd.read_parquet(TEST_FEAT_PATH)
return train, test
# LABELS_PATH = '../input/amex-default-prediction/train_labels.csv'
target = pd.read_csv(LABELS_PATH).target.values
train, test = get_data(read_from_cache=True)
print(f"target shape: {target.shape}, train shape: {train.shape}, test shape: {test.shape}")
target shape: (458913,), train shape: (458913, 469), test shape: (924621, 469)
# train columns 확인
features = [f for f in train.columns if f != 'customer_ID' and f != 'target']
features
['B_1_last', 'B_2_last', 'B_3_last', 'B_4_last', 'B_5_last', 'B_6_last', 'B_7_last', 'B_8_last', 'B_9_last', 'B_10_last', 'B_11_last', 'B_12_last', 'B_13_last', 'B_14_last', 'B_15_last', 'B_16_last', 'B_17_last', 'B_18_last', 'B_19_last', 'B_20_last', 'B_21_last', 'B_22_last', 'B_23_last', 'B_24_last', 'B_25_last', 'B_26_last', 'B_28_last', 'B_29_last', 'B_30_last', 'B_32_last', 'B_33_last', 'B_36_last', 'B_37_last', 'B_38_last', 'B_39_last', 'B_40_last', 'B_41_last', 'B_42_last', 'D_39_last', 'D_41_last', 'D_42_last', 'D_43_last', 'D_44_last', 'D_45_last', 'D_46_last', 'D_47_last', 'D_48_last', 'D_49_last', 'D_50_last', 'D_51_last', 'D_52_last', 'D_53_last', 'D_54_last', 'D_55_last', 'D_56_last', 'D_58_last', 'D_59_last', 'D_60_last', 'D_61_last', 'D_62_last', 'D_63_last', 'D_64_last', 'D_65_last', 'D_69_last', 'D_70_last', 'D_71_last', 'D_72_last', 'D_73_last', 'D_75_last', 'D_76_last', 'D_77_last', 'D_78_last', 'D_79_last', 'D_80_last', 'D_81_last', 'D_82_last', 'D_83_last', 'D_86_last', 'D_91_last', 'D_96_last', 'D_105_last', 'D_106_last', 'D_112_last', 'D_114_last', 'D_119_last', 'D_120_last', 'D_121_last', 'D_122_last', 'D_124_last', 'D_125_last', 'D_126_last', 'D_127_last', 'D_130_last', 'D_131_last', 'D_132_last', 'D_133_last', 'D_134_last', 'D_138_last', 'D_140_last', 'D_141_last', 'D_142_last', 'D_145_last', 'P_2_last', 'P_3_last', 'P_4_last', 'R_1_last', 'R_2_last', 'R_3_last', 'R_4_last', 'R_5_last', 'R_6_last', 'R_7_last', 'R_8_last', 'R_9_last', 'R_10_last', 'R_11_last', 'R_12_last', 'R_13_last', 'R_14_last', 'R_15_last', 'R_19_last', 'R_20_last', 'R_26_last', 'R_27_last', 'S_3_last', 'S_5_last', 'S_6_last', 'S_7_last', 'S_8_last', 'S_9_last', 'S_11_last', 'S_12_last', 'S_13_last', 'S_16_last', 'S_19_last', 'S_20_last', 'S_22_last', 'S_23_last', 'S_24_last', 'S_25_last', 'S_26_last', 'S_27_last', 'B_2_min', 'B_4_min', 'B_5_min', 'B_9_min', 'B_13_min', 'B_14_min', 'B_15_min', 'B_16_min', 'B_17_min', 'B_19_min', 'B_20_min', 'B_28_min', 'B_29_min', 'B_33_min', 'B_36_min', 'B_42_min', 'D_39_min', 'D_41_min', 'D_42_min', 'D_45_min', 'D_46_min', 'D_48_min', 'D_50_min', 'D_51_min', 'D_53_min', 'D_55_min', 'D_56_min', 'D_58_min', 'D_59_min', 'D_60_min', 'D_62_min', 'D_70_min', 'D_71_min', 'D_74_min', 'D_75_min', 'D_78_min', 'D_83_min', 'D_102_min', 'D_112_min', 'D_113_min', 'D_115_min', 'D_118_min', 'D_119_min', 'D_121_min', 'D_122_min', 'D_128_min', 'D_132_min', 'D_140_min', 'D_141_min', 'D_144_min', 'D_145_min', 'P_2_min', 'P_3_min', 'R_1_min', 'R_27_min', 'S_3_min', 'S_5_min', 'S_7_min', 'S_9_min', 'S_11_min', 'S_12_min', 'S_23_min', 'S_25_min', 'B_1_max', 'B_2_max', 'B_3_max', 'B_4_max', 'B_5_max', 'B_6_max', 'B_7_max', 'B_8_max', 'B_9_max', 'B_10_max', 'B_12_max', 'B_13_max', 'B_14_max', 'B_15_max', 'B_16_max', 'B_17_max', 'B_18_max', 'B_19_max', 'B_21_max', 'B_23_max', 'B_24_max', 'B_25_max', 'B_29_max', 'B_30_max', 'B_33_max', 'B_37_max', 'B_38_max', 'B_39_max', 'B_40_max', 'B_42_max', 'D_39_max', 'D_41_max', 'D_42_max', 'D_43_max', 'D_44_max', 'D_45_max', 'D_46_max', 'D_47_max', 'D_48_max', 'D_49_max', 'D_50_max', 'D_52_max', 'D_55_max', 'D_56_max', 'D_58_max', 'D_59_max', 'D_60_max', 'D_61_max', 'D_63_max', 'D_64_max', 'D_65_max', 'D_70_max', 'D_71_max', 'D_72_max', 'D_73_max', 'D_74_max', 'D_76_max', 'D_77_max', 'D_78_max', 'D_80_max', 'D_82_max', 'D_84_max', 'D_91_max', 'D_102_max', 'D_105_max', 'D_107_max', 'D_110_max', 'D_111_max', 'D_112_max', 'D_115_max', 'D_116_max', 'D_117_max', 'D_118_max', 'D_119_max', 'D_121_max', 'D_122_max', 'D_123_max', 'D_124_max', 'D_125_max', 'D_126_max', 'D_128_max', 'D_131_max', 'D_132_max', 'D_133_max', 'D_134_max', 'D_135_max', 'D_136_max', 'D_138_max', 'D_140_max', 'D_141_max', 'D_142_max', 'D_144_max', 'D_145_max', 'P_2_max', 'P_3_max', 'P_4_max', 'R_1_max', 'R_3_max', 'R_5_max', 'R_6_max', 'R_7_max', 'R_8_max', 'R_10_max', 'R_11_max', 'R_14_max', 'R_17_max', 'R_20_max', 'R_26_max', 'R_27_max', 'S_3_max', 'S_5_max', 'S_7_max', 'S_8_max', 'S_11_max', 'S_12_max', 'S_13_max', 'S_15_max', 'S_16_max', 'S_22_max', 'S_23_max', 'S_24_max', 'S_25_max', 'S_26_max', 'S_27_max', 'B_1_avg', 'B_2_avg', 'B_3_avg', 'B_4_avg', 'B_5_avg', 'B_6_avg', 'B_8_avg', 'B_9_avg', 'B_10_avg', 'B_11_avg', 'B_12_avg', 'B_13_avg', 'B_14_avg', 'B_15_avg', 'B_16_avg', 'B_17_avg', 'B_18_avg', 'B_19_avg', 'B_20_avg', 'B_21_avg', 'B_22_avg', 'B_23_avg', 'B_24_avg', 'B_25_avg', 'B_28_avg', 'B_29_avg', 'B_30_avg', 'B_32_avg', 'B_33_avg', 'B_37_avg', 'B_38_avg', 'B_39_avg', 'B_40_avg', 'B_41_avg', 'B_42_avg', 'D_39_avg', 'D_41_avg', 'D_42_avg', 'D_43_avg', 'D_44_avg', 'D_45_avg', 'D_46_avg', 'D_47_avg', 'D_48_avg', 'D_50_avg', 'D_51_avg', 'D_53_avg', 'D_54_avg', 'D_55_avg', 'D_58_avg', 'D_59_avg', 'D_60_avg', 'D_61_avg', 'D_62_avg', 'D_65_avg', 'D_66_avg', 'D_69_avg', 'D_70_avg', 'D_71_avg', 'D_72_avg', 'D_73_avg', 'D_74_avg', 'D_75_avg', 'D_76_avg', 'D_77_avg', 'D_78_avg', 'D_80_avg', 'D_82_avg', 'D_84_avg', 'D_86_avg', 'D_91_avg', 'D_92_avg', 'D_94_avg', 'D_96_avg', 'D_103_avg', 'D_104_avg', 'D_108_avg', 'D_112_avg', 'D_113_avg', 'D_114_avg', 'D_115_avg', 'D_117_avg', 'D_118_avg', 'D_119_avg', 'D_120_avg', 'D_121_avg', 'D_122_avg', 'D_123_avg', 'D_124_avg', 'D_125_avg', 'D_126_avg', 'D_128_avg', 'D_129_avg', 'D_131_avg', 'D_132_avg', 'D_133_avg', 'D_134_avg', 'D_135_avg', 'D_136_avg', 'D_140_avg', 'D_141_avg', 'D_142_avg', 'D_144_avg', 'D_145_avg', 'P_2_avg', 'P_3_avg', 'P_4_avg', 'R_1_avg', 'R_2_avg', 'R_3_avg', 'R_7_avg', 'R_8_avg', 'R_9_avg', 'R_10_avg', 'R_11_avg', 'R_14_avg', 'R_15_avg', 'R_16_avg', 'R_17_avg', 'R_20_avg', 'R_21_avg', 'R_22_avg', 'R_24_avg', 'R_26_avg', 'R_27_avg', 'S_3_avg', 'S_5_avg', 'S_6_avg', 'S_7_avg', 'S_9_avg', 'S_11_avg', 'S_12_avg', 'S_13_avg', 'S_15_avg', 'S_16_avg', 'S_18_avg', 'S_22_avg', 'S_23_avg', 'S_25_avg', 'S_26_avg']
param = {
"objective": "Logloss",
"learning_rate": 0.01,
"n_estimators": ITERATIONS,
#"eval_metric": "AUC", #AmexMetric(),
#'colsample_bylevel': 0.10506469029379303,
"max_depth": MAX_DEPTH,
#"l2_leaf_reg": 15,
"od_type": "Iter",
"od_wait": 600,
#"boosting_type": "Ordered",
#"bootstrap_type": "MVS",
"task_type":"GPU",
#"devices":'0:1',
#"auto_class_weights": "Balanced",
#"grow_policy": "Lossguide",
#"leaf_estimation_method": "Gradient",
#"leaf_estimation_iterations": 15,
#"leaf_estimation_backtracking": "Armijo",
"use_best_model": True,
#"scale_pos_weight": 20,
#"score_function": "L2"
}
cv_folds = 5
ONLY_FIRST_FOLD = False
score_list, y_pred_list = [], []
kf = StratifiedKFold(n_splits=cv_folds)
for fold, (idx_tr, idx_va) in enumerate(kf.split(train, target)):
X_tr, X_va, y_tr, y_va, model = None, None, None, None, None
start_time = datetime.datetime.now() # 시작 시간
X_tr = train.iloc[idx_tr][features]
X_va = train.iloc[idx_va][features]
y_tr = target[idx_tr]
y_va = target[idx_va]
with warnings.catch_warnings():
warnings.filterwarnings('ignore', category=UserWarning)
model = CatBoostClassifier(**param)
model.fit(X_tr, y_tr,
eval_set = [(X_va, y_va)],
metric_period=100
)
X_tr, y_tr = None, None
y_va_pred = model.predict_proba(X_va)[:,1]
score = amex_metric(y_va, y_va_pred)
n_trees = model.best_iteration_
if n_trees is None: n_trees = model.n_estimators
print(f"{Fore.GREEN}{Style.BRIGHT}Fold {fold} | {str(datetime.datetime.now() - start_time)[-12:-7]} |"
f" {n_trees:5} trees |"
f" Score = {score:.5f}{Style.RESET_ALL}")
score_list.append(score)
y_pred_list.append(model.predict_proba(test[features])[:,1])
gc.collect()
if ONLY_FIRST_FOLD: break # we only want the first fold
print(f"{Fore.GREEN}{Style.BRIGHT}OOF Score: {np.mean(score_list):.5f}{Style.RESET_ALL}")
gc.collect()
0: learn: 0.6791702 test: 0.6792409 best: 0.6792409 (0) total: 65.8ms remaining: 30m 43s 100: learn: 0.2710022 test: 0.2752627 best: 0.2752627 (100) total: 6.2s remaining: 28m 31s 200: learn: 0.2370489 test: 0.2438037 best: 0.2438037 (200) total: 12.8s remaining: 29m 27s 300: learn: 0.2269402 test: 0.2357646 best: 0.2357646 (300) total: 18.4s remaining: 28m 15s 400: learn: 0.2214221 test: 0.2321372 best: 0.2321372 (400) total: 24.4s remaining: 27m 59s 500: learn: 0.2173831 test: 0.2299194 best: 0.2299194 (500) total: 30s remaining: 27m 29s 600: learn: 0.2140748 test: 0.2284084 best: 0.2284084 (600) total: 36s remaining: 27m 23s 700: learn: 0.2111725 test: 0.2272325 best: 0.2272325 (700) total: 41.6s remaining: 27m 1s 800: learn: 0.2086000 test: 0.2263073 best: 0.2263073 (800) total: 48.2s remaining: 27m 17s 900: learn: 0.2062820 test: 0.2255653 best: 0.2255653 (900) total: 53.8s remaining: 26m 58s 1000: learn: 0.2041173 test: 0.2249418 best: 0.2249418 (1000) total: 59.8s remaining: 26m 51s 1100: learn: 0.2020286 test: 0.2244451 best: 0.2244451 (1100) total: 1m 5s remaining: 26m 35s 1200: learn: 0.2000949 test: 0.2240138 best: 0.2240138 (1200) total: 1m 11s remaining: 26m 27s 1300: learn: 0.1981670 test: 0.2235975 best: 0.2235975 (1300) total: 1m 16s remaining: 26m 13s 1400: learn: 0.1962536 test: 0.2232456 best: 0.2232456 (1400) total: 1m 23s remaining: 26m 20s 1500: learn: 0.1944423 test: 0.2229624 best: 0.2229624 (1500) total: 1m 29s remaining: 26m 13s 1600: learn: 0.1926103 test: 0.2226742 best: 0.2226742 (1600) total: 1m 34s remaining: 26m 1s 1700: learn: 0.1909520 test: 0.2224182 best: 0.2224182 (1700) total: 1m 40s remaining: 25m 53s 1800: learn: 0.1894082 test: 0.2221917 best: 0.2221904 (1799) total: 1m 45s remaining: 25m 40s 1900: learn: 0.1878429 test: 0.2219503 best: 0.2219503 (1900) total: 1m 51s remaining: 25m 35s 2000: learn: 0.1862920 test: 0.2217534 best: 0.2217534 (2000) total: 1m 57s remaining: 25m 32s 2100: learn: 0.1847936 test: 0.2215718 best: 0.2215718 (2100) total: 2m 3s remaining: 25m 25s 2200: learn: 0.1832632 test: 0.2213960 best: 0.2213960 (2200) total: 2m 9s remaining: 25m 15s 2300: learn: 0.1818048 test: 0.2212502 best: 0.2212502 (2300) total: 2m 15s remaining: 25m 8s 2400: learn: 0.1804099 test: 0.2211180 best: 0.2211180 (2400) total: 2m 20s remaining: 24m 58s 2500: learn: 0.1790014 test: 0.2209957 best: 0.2209957 (2500) total: 2m 27s remaining: 24m 59s 2600: learn: 0.1775942 test: 0.2208699 best: 0.2208697 (2599) total: 2m 32s remaining: 24m 50s 2700: learn: 0.1762968 test: 0.2207772 best: 0.2207772 (2700) total: 2m 38s remaining: 24m 44s 2800: learn: 0.1749489 test: 0.2206743 best: 0.2206741 (2799) total: 2m 43s remaining: 24m 34s 2900: learn: 0.1736309 test: 0.2205534 best: 0.2205534 (2900) total: 2m 49s remaining: 24m 29s 3000: learn: 0.1723057 test: 0.2204655 best: 0.2204655 (3000) total: 2m 55s remaining: 24m 20s 3100: learn: 0.1710175 test: 0.2203851 best: 0.2203851 (3100) total: 3m 1s remaining: 24m 19s 3200: learn: 0.1697786 test: 0.2202988 best: 0.2202979 (3197) total: 3m 7s remaining: 24m 11s 3300: learn: 0.1684702 test: 0.2202075 best: 0.2202070 (3298) total: 3m 12s remaining: 24m 4s 3400: learn: 0.1672348 test: 0.2201398 best: 0.2201364 (3394) total: 3m 18s remaining: 23m 58s 3500: learn: 0.1660494 test: 0.2200650 best: 0.2200650 (3500) total: 3m 24s remaining: 23m 49s 3600: learn: 0.1648015 test: 0.2200179 best: 0.2200153 (3589) total: 3m 30s remaining: 23m 43s 3700: learn: 0.1636077 test: 0.2199340 best: 0.2199320 (3695) total: 3m 36s remaining: 23m 39s 3800: learn: 0.1624856 test: 0.2198680 best: 0.2198680 (3800) total: 3m 42s remaining: 23m 33s 3900: learn: 0.1613351 test: 0.2197940 best: 0.2197940 (3900) total: 3m 47s remaining: 23m 25s 4000: learn: 0.1601564 test: 0.2197565 best: 0.2197519 (3991) total: 3m 53s remaining: 23m 19s 4100: learn: 0.1589886 test: 0.2197024 best: 0.2197003 (4099) total: 3m 58s remaining: 23m 11s 4200: learn: 0.1578875 test: 0.2196476 best: 0.2196476 (4200) total: 4m 5s remaining: 23m 9s 4300: learn: 0.1567437 test: 0.2196140 best: 0.2196130 (4298) total: 4m 10s remaining: 23m 1s 4400: learn: 0.1555321 test: 0.2195705 best: 0.2195705 (4400) total: 4m 16s remaining: 22m 56s 4500: learn: 0.1543614 test: 0.2195416 best: 0.2195415 (4499) total: 4m 22s remaining: 22m 48s 4600: learn: 0.1533140 test: 0.2194996 best: 0.2194987 (4597) total: 4m 27s remaining: 22m 42s 4700: learn: 0.1522279 test: 0.2194285 best: 0.2194282 (4696) total: 4m 33s remaining: 22m 34s 4800: learn: 0.1511772 test: 0.2193935 best: 0.2193935 (4800) total: 4m 39s remaining: 22m 32s 4900: learn: 0.1500883 test: 0.2193513 best: 0.2193503 (4898) total: 4m 45s remaining: 22m 24s 5000: learn: 0.1490020 test: 0.2193211 best: 0.2193202 (4999) total: 4m 51s remaining: 22m 18s 5100: learn: 0.1479950 test: 0.2192932 best: 0.2192918 (5099) total: 4m 56s remaining: 22m 11s 5200: learn: 0.1469498 test: 0.2192612 best: 0.2192595 (5197) total: 5m 2s remaining: 22m 5s 5300: learn: 0.1459204 test: 0.2192255 best: 0.2192234 (5291) total: 5m 8s remaining: 21m 58s 5400: learn: 0.1449383 test: 0.2191951 best: 0.2191951 (5400) total: 5m 14s remaining: 21m 55s 5500: learn: 0.1439535 test: 0.2191487 best: 0.2191481 (5499) total: 5m 20s remaining: 21m 49s 5600: learn: 0.1429408 test: 0.2191317 best: 0.2191299 (5598) total: 5m 25s remaining: 21m 42s 5700: learn: 0.1419854 test: 0.2191093 best: 0.2191015 (5674) total: 5m 31s remaining: 21m 36s 5800: learn: 0.1410402 test: 0.2190782 best: 0.2190768 (5796) total: 5m 36s remaining: 21m 29s 5900: learn: 0.1400868 test: 0.2190590 best: 0.2190561 (5895) total: 5m 42s remaining: 21m 23s 6000: learn: 0.1391050 test: 0.2190230 best: 0.2190224 (5998) total: 5m 48s remaining: 21m 19s 6100: learn: 0.1381035 test: 0.2190046 best: 0.2190046 (6100) total: 5m 54s remaining: 21m 13s 6200: learn: 0.1371542 test: 0.2189774 best: 0.2189774 (6200) total: 6m remaining: 21m 6s 6300: learn: 0.1361868 test: 0.2189528 best: 0.2189514 (6298) total: 6m 6s remaining: 21m 1s 6400: learn: 0.1352446 test: 0.2189250 best: 0.2189250 (6400) total: 6m 11s remaining: 20m 54s 6500: learn: 0.1343314 test: 0.2189129 best: 0.2189098 (6481) total: 6m 17s remaining: 20m 48s 6600: learn: 0.1334523 test: 0.2188849 best: 0.2188844 (6592) total: 6m 23s remaining: 20m 43s 6700: learn: 0.1325223 test: 0.2188869 best: 0.2188792 (6626) total: 6m 29s remaining: 20m 38s 6800: learn: 0.1316275 test: 0.2188684 best: 0.2188665 (6797) total: 6m 34s remaining: 20m 31s 6900: learn: 0.1307516 test: 0.2188505 best: 0.2188471 (6876) total: 6m 40s remaining: 20m 25s 7000: learn: 0.1298031 test: 0.2188339 best: 0.2188331 (6996) total: 6m 46s remaining: 20m 18s 7100: learn: 0.1289074 test: 0.2188105 best: 0.2188105 (7100) total: 6m 52s remaining: 20m 14s 7200: learn: 0.1280325 test: 0.2187762 best: 0.2187737 (7195) total: 6m 58s remaining: 20m 9s 7300: learn: 0.1272251 test: 0.2187899 best: 0.2187737 (7195) total: 7m 4s remaining: 20m 2s 7400: learn: 0.1263482 test: 0.2187888 best: 0.2187737 (7195) total: 7m 9s remaining: 19m 56s 7500: learn: 0.1254896 test: 0.2187712 best: 0.2187692 (7495) total: 7m 15s remaining: 19m 49s 7600: learn: 0.1246411 test: 0.2187693 best: 0.2187643 (7566) total: 7m 21s remaining: 19m 44s 7700: learn: 0.1238167 test: 0.2187275 best: 0.2187244 (7697) total: 7m 27s remaining: 19m 39s 7800: learn: 0.1229858 test: 0.2187211 best: 0.2187200 (7799) total: 7m 33s remaining: 19m 33s 7900: learn: 0.1222194 test: 0.2187047 best: 0.2187036 (7894) total: 7m 38s remaining: 19m 27s 8000: learn: 0.1213965 test: 0.2187065 best: 0.2186995 (7947) total: 7m 44s remaining: 19m 21s 8100: learn: 0.1205898 test: 0.2187167 best: 0.2186995 (7947) total: 7m 50s remaining: 19m 14s 8200: learn: 0.1197957 test: 0.2187125 best: 0.2186995 (7947) total: 7m 55s remaining: 19m 8s 8300: learn: 0.1190082 test: 0.2186945 best: 0.2186905 (8284) total: 8m 2s remaining: 19m 4s 8400: learn: 0.1182592 test: 0.2186896 best: 0.2186890 (8374) total: 8m 7s remaining: 18m 58s 8500: learn: 0.1174451 test: 0.2186695 best: 0.2186658 (8491) total: 8m 13s remaining: 18m 51s 8600: learn: 0.1166784 test: 0.2186721 best: 0.2186620 (8529) total: 8m 19s remaining: 18m 45s 8700: learn: 0.1159210 test: 0.2186711 best: 0.2186620 (8529) total: 8m 24s remaining: 18m 39s 8800: learn: 0.1151929 test: 0.2186705 best: 0.2186620 (8529) total: 8m 30s remaining: 18m 33s 8900: learn: 0.1144817 test: 0.2186585 best: 0.2186585 (8900) total: 8m 36s remaining: 18m 28s 9000: learn: 0.1137264 test: 0.2186636 best: 0.2186490 (8931) total: 8m 42s remaining: 18m 22s 9100: learn: 0.1130065 test: 0.2186711 best: 0.2186490 (8931) total: 8m 48s remaining: 18m 17s 9200: learn: 0.1122937 test: 0.2186741 best: 0.2186490 (8931) total: 8m 53s remaining: 18m 10s 9300: learn: 0.1115545 test: 0.2186652 best: 0.2186490 (8931) total: 8m 59s remaining: 18m 4s 9400: learn: 0.1108159 test: 0.2186588 best: 0.2186490 (8931) total: 9m 5s remaining: 17m 59s 9500: learn: 0.1100934 test: 0.2186622 best: 0.2186490 (8931) total: 9m 11s remaining: 17m 53s bestTest = 0.2186489635 bestIteration = 8931 Shrink model to first 8932 iterations. Fold 0 | 10:52 | 8931 trees | Score = 0.79373 0: learn: 0.6791823 test: 0.6792041 best: 0.6792041 (0) total: 61.7ms remaining: 28m 48s 100: learn: 0.2711499 test: 0.2743723 best: 0.2743723 (100) total: 6.9s remaining: 31m 44s 200: learn: 0.2372537 test: 0.2427864 best: 0.2427864 (200) total: 12.6s remaining: 28m 59s 300: learn: 0.2272944 test: 0.2347313 best: 0.2347313 (300) total: 18.6s remaining: 28m 28s 400: learn: 0.2218055 test: 0.2310920 best: 0.2310920 (400) total: 24.3s remaining: 27m 51s 500: learn: 0.2177323 test: 0.2288029 best: 0.2288029 (500) total: 30.2s remaining: 27m 36s 600: learn: 0.2145320 test: 0.2272759 best: 0.2272759 (600) total: 36.1s remaining: 27m 27s 700: learn: 0.2116936 test: 0.2261204 best: 0.2261204 (700) total: 41.7s remaining: 27m 3s 800: learn: 0.2091639 test: 0.2252335 best: 0.2252335 (800) total: 48.2s remaining: 27m 16s 900: learn: 0.2068577 test: 0.2245083 best: 0.2245083 (900) total: 53.7s remaining: 26m 55s 1000: learn: 0.2047223 test: 0.2239207 best: 0.2239207 (1000) total: 59.6s remaining: 26m 47s 1100: learn: 0.2026296 test: 0.2233980 best: 0.2233980 (1100) total: 1m 5s remaining: 26m 32s 1200: learn: 0.2006089 test: 0.2229966 best: 0.2229966 (1200) total: 1m 11s remaining: 26m 25s 1300: learn: 0.1987032 test: 0.2226086 best: 0.2226086 (1300) total: 1m 16s remaining: 26m 10s 1400: learn: 0.1967927 test: 0.2222908 best: 0.2222908 (1400) total: 1m 23s remaining: 26m 18s 1500: learn: 0.1949216 test: 0.2219832 best: 0.2219832 (1500) total: 1m 28s remaining: 26m 5s 1600: learn: 0.1931617 test: 0.2217173 best: 0.2217173 (1600) total: 1m 34s remaining: 25m 58s 1700: learn: 0.1914876 test: 0.2214893 best: 0.2214893 (1700) total: 1m 40s remaining: 25m 46s 1800: learn: 0.1898755 test: 0.2212723 best: 0.2212723 (1800) total: 1m 45s remaining: 25m 40s 1900: learn: 0.1883055 test: 0.2210956 best: 0.2210956 (1899) total: 1m 52s remaining: 25m 40s 2000: learn: 0.1867360 test: 0.2209228 best: 0.2209228 (2000) total: 1m 58s remaining: 25m 33s 2100: learn: 0.1852174 test: 0.2207493 best: 0.2207493 (2100) total: 2m 3s remaining: 25m 24s 2200: learn: 0.1837739 test: 0.2206249 best: 0.2206242 (2199) total: 2m 9s remaining: 25m 15s 2300: learn: 0.1823122 test: 0.2204892 best: 0.2204892 (2300) total: 2m 15s remaining: 25m 8s 2400: learn: 0.1808832 test: 0.2203595 best: 0.2203595 (2400) total: 2m 20s remaining: 24m 57s 2500: learn: 0.1794797 test: 0.2202550 best: 0.2202550 (2500) total: 2m 27s remaining: 24m 58s 2600: learn: 0.1781280 test: 0.2201227 best: 0.2201217 (2599) total: 2m 32s remaining: 24m 48s 2700: learn: 0.1767615 test: 0.2200097 best: 0.2200088 (2698) total: 2m 38s remaining: 24m 42s 2800: learn: 0.1753805 test: 0.2199180 best: 0.2199180 (2800) total: 2m 43s remaining: 24m 33s 2900: learn: 0.1740182 test: 0.2198225 best: 0.2198225 (2900) total: 2m 49s remaining: 24m 27s 3000: learn: 0.1726605 test: 0.2197172 best: 0.2197172 (3000) total: 2m 55s remaining: 24m 18s 3100: learn: 0.1714069 test: 0.2196520 best: 0.2196503 (3097) total: 3m 1s remaining: 24m 17s 3200: learn: 0.1701806 test: 0.2195665 best: 0.2195665 (3200) total: 3m 6s remaining: 24m 8s 3300: learn: 0.1688418 test: 0.2194812 best: 0.2194812 (3300) total: 3m 12s remaining: 24m 2s 3400: learn: 0.1675889 test: 0.2194199 best: 0.2194173 (3399) total: 3m 18s remaining: 23m 53s 3500: learn: 0.1662976 test: 0.2193283 best: 0.2193283 (3500) total: 3m 24s remaining: 23m 48s 3600: learn: 0.1650171 test: 0.2192620 best: 0.2192620 (3600) total: 3m 30s remaining: 23m 45s 3700: learn: 0.1638330 test: 0.2192054 best: 0.2192047 (3698) total: 3m 36s remaining: 23m 40s 3800: learn: 0.1625818 test: 0.2191579 best: 0.2191578 (3798) total: 3m 41s remaining: 23m 32s 3900: learn: 0.1613933 test: 0.2191016 best: 0.2191016 (3900) total: 3m 47s remaining: 23m 27s 4000: learn: 0.1601823 test: 0.2190454 best: 0.2190454 (4000) total: 3m 53s remaining: 23m 19s 4100: learn: 0.1589788 test: 0.2190098 best: 0.2190066 (4069) total: 3m 59s remaining: 23m 13s 4200: learn: 0.1578199 test: 0.2189614 best: 0.2189614 (4200) total: 4m 5s remaining: 23m 11s 4300: learn: 0.1566920 test: 0.2189087 best: 0.2189072 (4291) total: 4m 11s remaining: 23m 3s 4400: learn: 0.1554957 test: 0.2188704 best: 0.2188704 (4400) total: 4m 16s remaining: 22m 57s 4500: learn: 0.1544380 test: 0.2188288 best: 0.2188288 (4500) total: 4m 22s remaining: 22m 50s 4600: learn: 0.1533411 test: 0.2188148 best: 0.2188144 (4599) total: 4m 28s remaining: 22m 44s 4700: learn: 0.1522799 test: 0.2187715 best: 0.2187710 (4699) total: 4m 33s remaining: 22m 37s 4800: learn: 0.1512079 test: 0.2187466 best: 0.2187416 (4781) total: 4m 40s remaining: 22m 34s 4900: learn: 0.1501659 test: 0.2187174 best: 0.2187157 (4897) total: 4m 45s remaining: 22m 26s 5000: learn: 0.1491330 test: 0.2186934 best: 0.2186934 (5000) total: 4m 51s remaining: 22m 20s 5100: learn: 0.1480773 test: 0.2186630 best: 0.2186620 (5088) total: 4m 57s remaining: 22m 13s 5200: learn: 0.1470362 test: 0.2186382 best: 0.2186382 (5200) total: 5m 2s remaining: 22m 7s 5300: learn: 0.1459535 test: 0.2185877 best: 0.2185856 (5293) total: 5m 9s remaining: 22m 3s 5400: learn: 0.1449251 test: 0.2185558 best: 0.2185558 (5400) total: 5m 14s remaining: 21m 57s 5500: learn: 0.1440058 test: 0.2185297 best: 0.2185297 (5500) total: 5m 20s remaining: 21m 49s 5600: learn: 0.1429900 test: 0.2185038 best: 0.2185038 (5600) total: 5m 26s remaining: 21m 44s 5700: learn: 0.1419718 test: 0.2184832 best: 0.2184818 (5699) total: 5m 31s remaining: 21m 37s 5800: learn: 0.1410126 test: 0.2184575 best: 0.2184564 (5798) total: 5m 37s remaining: 21m 31s 5900: learn: 0.1400430 test: 0.2184419 best: 0.2184407 (5894) total: 5m 43s remaining: 21m 27s 6000: learn: 0.1390600 test: 0.2184212 best: 0.2184194 (5998) total: 5m 49s remaining: 21m 20s 6100: learn: 0.1381357 test: 0.2184108 best: 0.2184077 (6064) total: 5m 55s remaining: 21m 14s 6200: learn: 0.1371843 test: 0.2183879 best: 0.2183834 (6195) total: 6m remaining: 21m 7s 6300: learn: 0.1362755 test: 0.2183726 best: 0.2183679 (6278) total: 6m 6s remaining: 21m 1s 6400: learn: 0.1352992 test: 0.2183569 best: 0.2183558 (6394) total: 6m 11s remaining: 20m 54s 6500: learn: 0.1343295 test: 0.2183587 best: 0.2183558 (6394) total: 6m 18s remaining: 20m 51s 6600: learn: 0.1333872 test: 0.2183460 best: 0.2183399 (6579) total: 6m 23s remaining: 20m 44s 6700: learn: 0.1324633 test: 0.2183206 best: 0.2183194 (6687) total: 6m 29s remaining: 20m 38s 6800: learn: 0.1316374 test: 0.2182789 best: 0.2182789 (6798) total: 6m 35s remaining: 20m 31s 6900: learn: 0.1307106 test: 0.2182753 best: 0.2182676 (6832) total: 6m 41s remaining: 20m 26s 7000: learn: 0.1298194 test: 0.2182624 best: 0.2182559 (6972) total: 6m 47s remaining: 20m 21s 7100: learn: 0.1289649 test: 0.2182616 best: 0.2182559 (6972) total: 6m 53s remaining: 20m 15s 7200: learn: 0.1280979 test: 0.2182495 best: 0.2182495 (7200) total: 6m 58s remaining: 20m 8s 7300: learn: 0.1272505 test: 0.2182413 best: 0.2182413 (7300) total: 7m 4s remaining: 20m 3s 7400: learn: 0.1263786 test: 0.2182375 best: 0.2182356 (7398) total: 7m 9s remaining: 19m 56s 7500: learn: 0.1255040 test: 0.2182310 best: 0.2182251 (7472) total: 7m 15s remaining: 19m 50s 7600: learn: 0.1246796 test: 0.2182285 best: 0.2182158 (7542) total: 7m 21s remaining: 19m 45s 7700: learn: 0.1238437 test: 0.2182236 best: 0.2182158 (7542) total: 7m 27s remaining: 19m 39s 7800: learn: 0.1230118 test: 0.2182245 best: 0.2182158 (7542) total: 7m 33s remaining: 19m 33s 7900: learn: 0.1221879 test: 0.2182263 best: 0.2182158 (7542) total: 7m 38s remaining: 19m 27s 8000: learn: 0.1213753 test: 0.2182195 best: 0.2182117 (7988) total: 7m 44s remaining: 19m 21s 8100: learn: 0.1205973 test: 0.2182204 best: 0.2182117 (7988) total: 7m 50s remaining: 19m 14s 8200: learn: 0.1198612 test: 0.2182152 best: 0.2182079 (8119) total: 7m 56s remaining: 19m 10s 8300: learn: 0.1190333 test: 0.2182091 best: 0.2182079 (8119) total: 8m 2s remaining: 19m 4s 8400: learn: 0.1182459 test: 0.2182108 best: 0.2182079 (8119) total: 8m 7s remaining: 18m 58s 8500: learn: 0.1174931 test: 0.2182179 best: 0.2182020 (8452) total: 8m 13s remaining: 18m 51s 8600: learn: 0.1167373 test: 0.2182294 best: 0.2182020 (8452) total: 8m 19s remaining: 18m 45s 8700: learn: 0.1159673 test: 0.2182250 best: 0.2182020 (8452) total: 8m 25s remaining: 18m 40s 8800: learn: 0.1152444 test: 0.2182408 best: 0.2182020 (8452) total: 8m 31s remaining: 18m 34s 8900: learn: 0.1144861 test: 0.2182503 best: 0.2182020 (8452) total: 8m 36s remaining: 18m 28s 9000: learn: 0.1137235 test: 0.2182581 best: 0.2182020 (8452) total: 8m 42s remaining: 18m 22s bestTest = 0.2182020236 bestIteration = 8452 Shrink model to first 8453 iterations. Fold 1 | 09:01 | 8452 trees | Score = 0.79032 0: learn: 0.6792632 test: 0.6792852 best: 0.6792852 (0) total: 66.6ms remaining: 31m 5s 100: learn: 0.2715080 test: 0.2733500 best: 0.2733500 (100) total: 6.11s remaining: 28m 7s 200: learn: 0.2377051 test: 0.2412989 best: 0.2412989 (200) total: 11.8s remaining: 27m 5s 300: learn: 0.2275879 test: 0.2328313 best: 0.2328313 (300) total: 17.6s remaining: 27m 2s 400: learn: 0.2221463 test: 0.2290515 best: 0.2290515 (400) total: 23.9s remaining: 27m 22s 500: learn: 0.2180765 test: 0.2267600 best: 0.2267600 (500) total: 29.7s remaining: 27m 9s 600: learn: 0.2149154 test: 0.2252118 best: 0.2252118 (600) total: 35.4s remaining: 26m 55s 700: learn: 0.2119658 test: 0.2240597 best: 0.2240597 (700) total: 41.1s remaining: 26m 40s 800: learn: 0.2094026 test: 0.2231804 best: 0.2231804 (800) total: 46.9s remaining: 26m 31s 900: learn: 0.2070280 test: 0.2224819 best: 0.2224819 (900) total: 52.6s remaining: 26m 22s 1000: learn: 0.2048516 test: 0.2218968 best: 0.2218968 (1000) total: 59s remaining: 26m 31s 1100: learn: 0.2027083 test: 0.2213747 best: 0.2213747 (1100) total: 1m 4s remaining: 26m 16s 1200: learn: 0.2007579 test: 0.2209609 best: 0.2209609 (1200) total: 1m 10s remaining: 26m 10s 1300: learn: 0.1987487 test: 0.2206191 best: 0.2206191 (1300) total: 1m 15s remaining: 25m 57s 1400: learn: 0.1968963 test: 0.2203039 best: 0.2203039 (1400) total: 1m 21s remaining: 25m 49s 1500: learn: 0.1950406 test: 0.2200247 best: 0.2200247 (1500) total: 1m 27s remaining: 25m 37s 1600: learn: 0.1932656 test: 0.2198108 best: 0.2198108 (1600) total: 1m 33s remaining: 25m 42s 1700: learn: 0.1915794 test: 0.2196299 best: 0.2196292 (1697) total: 1m 38s remaining: 25m 30s 1800: learn: 0.1899830 test: 0.2194281 best: 0.2194281 (1800) total: 1m 44s remaining: 25m 23s 1900: learn: 0.1884027 test: 0.2192661 best: 0.2192661 (1900) total: 1m 50s remaining: 25m 12s 2000: learn: 0.1868138 test: 0.2191302 best: 0.2191302 (2000) total: 1m 55s remaining: 25m 6s 2100: learn: 0.1853534 test: 0.2189799 best: 0.2189799 (2100) total: 2m 1s remaining: 24m 56s 2200: learn: 0.1838680 test: 0.2188367 best: 0.2188367 (2200) total: 2m 7s remaining: 24m 56s 2300: learn: 0.1824095 test: 0.2187172 best: 0.2187172 (2300) total: 2m 13s remaining: 24m 47s 2400: learn: 0.1809152 test: 0.2185895 best: 0.2185895 (2400) total: 2m 18s remaining: 24m 40s 2500: learn: 0.1795447 test: 0.2184939 best: 0.2184937 (2497) total: 2m 24s remaining: 24m 31s 2600: learn: 0.1780811 test: 0.2183817 best: 0.2183817 (2600) total: 2m 30s remaining: 24m 25s 2700: learn: 0.1767149 test: 0.2182694 best: 0.2182694 (2700) total: 2m 35s remaining: 24m 15s 2800: learn: 0.1753583 test: 0.2181993 best: 0.2181993 (2800) total: 2m 41s remaining: 24m 14s 2900: learn: 0.1740377 test: 0.2181147 best: 0.2181147 (2900) total: 2m 47s remaining: 24m 6s 3000: learn: 0.1726672 test: 0.2180209 best: 0.2180205 (2999) total: 2m 52s remaining: 23m 59s 3100: learn: 0.1713258 test: 0.2179355 best: 0.2179355 (3100) total: 2m 58s remaining: 23m 53s 3200: learn: 0.1700902 test: 0.2178752 best: 0.2178752 (3200) total: 3m 3s remaining: 23m 45s 3300: learn: 0.1688247 test: 0.2178143 best: 0.2178143 (3300) total: 3m 9s remaining: 23m 39s 3400: learn: 0.1675882 test: 0.2177431 best: 0.2177416 (3395) total: 3m 15s remaining: 23m 36s 3500: learn: 0.1662843 test: 0.2176941 best: 0.2176941 (3500) total: 3m 21s remaining: 23m 30s 3600: learn: 0.1650635 test: 0.2176228 best: 0.2176228 (3600) total: 3m 27s remaining: 23m 22s 3700: learn: 0.1638443 test: 0.2175643 best: 0.2175623 (3697) total: 3m 32s remaining: 23m 16s 3800: learn: 0.1626021 test: 0.2175333 best: 0.2175324 (3792) total: 3m 38s remaining: 23m 9s 3900: learn: 0.1614136 test: 0.2174943 best: 0.2174943 (3900) total: 3m 43s remaining: 23m 3s 4000: learn: 0.1602729 test: 0.2174429 best: 0.2174414 (3981) total: 3m 49s remaining: 22m 59s 4100: learn: 0.1591683 test: 0.2174060 best: 0.2174048 (4097) total: 3m 55s remaining: 22m 53s 4200: learn: 0.1579892 test: 0.2173624 best: 0.2173605 (4199) total: 4m 1s remaining: 22m 45s 4300: learn: 0.1568576 test: 0.2173169 best: 0.2173131 (4299) total: 4m 6s remaining: 22m 39s 4400: learn: 0.1557020 test: 0.2172625 best: 0.2172622 (4399) total: 4m 12s remaining: 22m 32s 4500: learn: 0.1546586 test: 0.2172393 best: 0.2172385 (4498) total: 4m 17s remaining: 22m 26s 4600: learn: 0.1535849 test: 0.2172078 best: 0.2172060 (4563) total: 4m 24s remaining: 22m 23s 4700: learn: 0.1524028 test: 0.2171542 best: 0.2171532 (4699) total: 4m 29s remaining: 22m 17s 4800: learn: 0.1513169 test: 0.2171393 best: 0.2171390 (4790) total: 4m 35s remaining: 22m 10s 4900: learn: 0.1502911 test: 0.2171279 best: 0.2171243 (4893) total: 4m 41s remaining: 22m 4s 5000: learn: 0.1492534 test: 0.2171117 best: 0.2171046 (4981) total: 4m 46s remaining: 21m 57s 5100: learn: 0.1482265 test: 0.2170955 best: 0.2170943 (5083) total: 4m 52s remaining: 21m 51s 5200: learn: 0.1472342 test: 0.2170581 best: 0.2170581 (5200) total: 4m 58s remaining: 21m 47s 5300: learn: 0.1462111 test: 0.2170267 best: 0.2170248 (5283) total: 5m 3s remaining: 21m 41s 5400: learn: 0.1451768 test: 0.2170054 best: 0.2170022 (5395) total: 5m 9s remaining: 21m 34s 5500: learn: 0.1441352 test: 0.2170074 best: 0.2170009 (5460) total: 5m 15s remaining: 21m 28s 5600: learn: 0.1431382 test: 0.2169735 best: 0.2169716 (5596) total: 5m 20s remaining: 21m 21s 5700: learn: 0.1421954 test: 0.2169444 best: 0.2169436 (5695) total: 5m 26s remaining: 21m 15s 5800: learn: 0.1411624 test: 0.2169207 best: 0.2169200 (5797) total: 5m 32s remaining: 21m 13s 5900: learn: 0.1401568 test: 0.2169144 best: 0.2169106 (5867) total: 5m 38s remaining: 21m 6s 6000: learn: 0.1391751 test: 0.2168876 best: 0.2168876 (6000) total: 5m 43s remaining: 21m 6100: learn: 0.1381957 test: 0.2168552 best: 0.2168515 (6094) total: 5m 49s remaining: 20m 53s 6200: learn: 0.1371681 test: 0.2168271 best: 0.2168259 (6199) total: 5m 55s remaining: 20m 48s 6300: learn: 0.1363106 test: 0.2168196 best: 0.2168196 (6300) total: 6m remaining: 20m 41s 6400: learn: 0.1353621 test: 0.2168076 best: 0.2168066 (6395) total: 6m 6s remaining: 20m 35s 6500: learn: 0.1344492 test: 0.2167932 best: 0.2167905 (6499) total: 6m 12s remaining: 20m 31s 6600: learn: 0.1335427 test: 0.2167820 best: 0.2167816 (6599) total: 6m 17s remaining: 20m 25s 6700: learn: 0.1325955 test: 0.2167656 best: 0.2167650 (6698) total: 6m 23s remaining: 20m 18s 6800: learn: 0.1317129 test: 0.2167663 best: 0.2167628 (6791) total: 6m 29s remaining: 20m 13s 6900: learn: 0.1308030 test: 0.2167724 best: 0.2167554 (6855) total: 6m 34s remaining: 20m 6s 7000: learn: 0.1299023 test: 0.2167563 best: 0.2167524 (6994) total: 6m 40s remaining: 20m 7100: learn: 0.1290363 test: 0.2167597 best: 0.2167439 (7036) total: 6m 46s remaining: 19m 56s 7200: learn: 0.1281465 test: 0.2167592 best: 0.2167439 (7036) total: 6m 52s remaining: 19m 50s 7300: learn: 0.1272759 test: 0.2167697 best: 0.2167439 (7036) total: 6m 57s remaining: 19m 44s 7400: learn: 0.1264581 test: 0.2167717 best: 0.2167439 (7036) total: 7m 3s remaining: 19m 38s 7500: learn: 0.1256158 test: 0.2167902 best: 0.2167439 (7036) total: 7m 8s remaining: 19m 32s 7600: learn: 0.1247645 test: 0.2167765 best: 0.2167439 (7036) total: 7m 14s remaining: 19m 26s bestTest = 0.2167438883 bestIteration = 7036 Shrink model to first 7037 iterations. Fold 2 | 07:31 | 7036 trees | Score = 0.79282 0: learn: 0.6794135 test: 0.6794286 best: 0.6794286 (0) total: 62.6ms remaining: 29m 13s 100: learn: 0.2718769 test: 0.2732579 best: 0.2732579 (100) total: 6.17s remaining: 28m 25s 200: learn: 0.2380409 test: 0.2406725 best: 0.2406725 (200) total: 12.2s remaining: 28m 4s 300: learn: 0.2280781 test: 0.2321630 best: 0.2321630 (300) total: 17.8s remaining: 27m 19s 400: learn: 0.2225044 test: 0.2282960 best: 0.2282960 (400) total: 24.5s remaining: 28m 9s 500: learn: 0.2184056 test: 0.2259073 best: 0.2259073 (500) total: 30.2s remaining: 27m 35s 600: learn: 0.2151412 test: 0.2243272 best: 0.2243272 (600) total: 36.1s remaining: 27m 25s 700: learn: 0.2123174 test: 0.2231223 best: 0.2231223 (700) total: 41.6s remaining: 27m 1s 800: learn: 0.2097466 test: 0.2221943 best: 0.2221943 (800) total: 47.5s remaining: 26m 53s 900: learn: 0.2073053 test: 0.2214205 best: 0.2214205 (900) total: 53.1s remaining: 26m 36s 1000: learn: 0.2051111 test: 0.2207890 best: 0.2207890 (1000) total: 59.2s remaining: 26m 36s 1100: learn: 0.2029648 test: 0.2202731 best: 0.2202731 (1100) total: 1m 5s remaining: 26m 32s 1200: learn: 0.2009483 test: 0.2198401 best: 0.2198401 (1200) total: 1m 11s remaining: 26m 24s 1300: learn: 0.1989643 test: 0.2194413 best: 0.2194413 (1300) total: 1m 16s remaining: 26m 11s 1400: learn: 0.1970485 test: 0.2190814 best: 0.2190814 (1400) total: 1m 22s remaining: 26m 4s 1500: learn: 0.1951877 test: 0.2187517 best: 0.2187517 (1500) total: 1m 27s remaining: 25m 52s 1600: learn: 0.1934322 test: 0.2185053 best: 0.2185053 (1600) total: 1m 34s remaining: 25m 59s 1700: learn: 0.1917309 test: 0.2182963 best: 0.2182963 (1700) total: 1m 40s remaining: 25m 52s 1800: learn: 0.1901457 test: 0.2180993 best: 0.2180993 (1800) total: 1m 45s remaining: 25m 40s 1900: learn: 0.1885070 test: 0.2179172 best: 0.2179172 (1900) total: 1m 51s remaining: 25m 33s 2000: learn: 0.1869324 test: 0.2177336 best: 0.2177336 (2000) total: 1m 57s remaining: 25m 22s 2100: learn: 0.1853162 test: 0.2175839 best: 0.2175839 (2100) total: 2m 3s remaining: 25m 16s 2200: learn: 0.1838502 test: 0.2174315 best: 0.2174315 (2200) total: 2m 8s remaining: 25m 6s 2300: learn: 0.1823480 test: 0.2172696 best: 0.2172696 (2300) total: 2m 14s remaining: 25m 6s 2400: learn: 0.1809113 test: 0.2171429 best: 0.2171429 (2400) total: 2m 20s remaining: 24m 56s 2500: learn: 0.1794845 test: 0.2170374 best: 0.2170374 (2499) total: 2m 26s remaining: 24m 50s 2600: learn: 0.1781053 test: 0.2169134 best: 0.2169134 (2600) total: 2m 31s remaining: 24m 40s 2700: learn: 0.1766701 test: 0.2168166 best: 0.2168166 (2700) total: 2m 37s remaining: 24m 34s 2800: learn: 0.1752213 test: 0.2167102 best: 0.2167102 (2800) total: 2m 42s remaining: 24m 25s 2900: learn: 0.1738758 test: 0.2166150 best: 0.2166150 (2900) total: 2m 48s remaining: 24m 19s 3000: learn: 0.1725159 test: 0.2165121 best: 0.2165120 (2999) total: 2m 54s remaining: 24m 15s 3100: learn: 0.1711763 test: 0.2164259 best: 0.2164259 (3100) total: 3m remaining: 24m 9s 3200: learn: 0.1698825 test: 0.2163755 best: 0.2163713 (3176) total: 3m 5s remaining: 24m 3300: learn: 0.1685658 test: 0.2163266 best: 0.2163266 (3300) total: 3m 11s remaining: 23m 54s 3400: learn: 0.1673194 test: 0.2162499 best: 0.2162499 (3400) total: 3m 17s remaining: 23m 45s 3500: learn: 0.1660572 test: 0.2161687 best: 0.2161687 (3500) total: 3m 22s remaining: 23m 39s 3600: learn: 0.1648510 test: 0.2160975 best: 0.2160975 (3599) total: 3m 28s remaining: 23m 35s 3700: learn: 0.1635932 test: 0.2160186 best: 0.2160146 (3696) total: 3m 34s remaining: 23m 30s 3800: learn: 0.1624251 test: 0.2159669 best: 0.2159669 (3800) total: 3m 40s remaining: 23m 23s 3900: learn: 0.1612134 test: 0.2159154 best: 0.2159144 (3898) total: 3m 46s remaining: 23m 17s 4000: learn: 0.1600452 test: 0.2158746 best: 0.2158732 (3996) total: 3m 51s remaining: 23m 11s 4100: learn: 0.1588433 test: 0.2158113 best: 0.2158113 (4100) total: 3m 57s remaining: 23m 4s 4200: learn: 0.1576475 test: 0.2157576 best: 0.2157536 (4198) total: 4m 4s remaining: 23m 2s 4300: learn: 0.1564285 test: 0.2157251 best: 0.2157245 (4296) total: 4m 9s remaining: 22m 55s 4400: learn: 0.1552672 test: 0.2156958 best: 0.2156958 (4400) total: 4m 15s remaining: 22m 49s 4500: learn: 0.1542060 test: 0.2156514 best: 0.2156514 (4497) total: 4m 20s remaining: 22m 41s 4600: learn: 0.1531216 test: 0.2156204 best: 0.2156177 (4592) total: 4m 26s remaining: 22m 35s 4700: learn: 0.1520646 test: 0.2155714 best: 0.2155714 (4700) total: 4m 32s remaining: 22m 31s 4800: learn: 0.1509923 test: 0.2155295 best: 0.2155245 (4795) total: 4m 38s remaining: 22m 25s 4900: learn: 0.1499344 test: 0.2154834 best: 0.2154822 (4898) total: 4m 44s remaining: 22m 18s 5000: learn: 0.1489365 test: 0.2154531 best: 0.2154529 (4999) total: 4m 49s remaining: 22m 12s 5100: learn: 0.1478536 test: 0.2154076 best: 0.2154074 (5099) total: 4m 55s remaining: 22m 5s 5200: learn: 0.1468287 test: 0.2153692 best: 0.2153685 (5196) total: 5m 1s remaining: 22m 5300: learn: 0.1458253 test: 0.2153385 best: 0.2153385 (5300) total: 5m 7s remaining: 21m 55s 5400: learn: 0.1447986 test: 0.2153220 best: 0.2153220 (5400) total: 5m 13s remaining: 21m 49s 5500: learn: 0.1438051 test: 0.2152830 best: 0.2152830 (5500) total: 5m 18s remaining: 21m 42s 5600: learn: 0.1427821 test: 0.2152513 best: 0.2152508 (5597) total: 5m 24s remaining: 21m 36s 5700: learn: 0.1417825 test: 0.2152212 best: 0.2152159 (5676) total: 5m 29s remaining: 21m 29s 5800: learn: 0.1407505 test: 0.2151993 best: 0.2151993 (5800) total: 5m 35s remaining: 21m 24s 5900: learn: 0.1397435 test: 0.2151756 best: 0.2151741 (5898) total: 5m 41s remaining: 21m 17s 6000: learn: 0.1387245 test: 0.2151609 best: 0.2151573 (5990) total: 5m 46s remaining: 21m 11s 6100: learn: 0.1377645 test: 0.2151372 best: 0.2151337 (6099) total: 5m 52s remaining: 21m 6s 6200: learn: 0.1368441 test: 0.2151241 best: 0.2151215 (6161) total: 5m 58s remaining: 21m 1s 6300: learn: 0.1359154 test: 0.2151299 best: 0.2151215 (6161) total: 6m 4s remaining: 20m 55s 6400: learn: 0.1349763 test: 0.2151087 best: 0.2151061 (6372) total: 6m 10s remaining: 20m 49s 6500: learn: 0.1340231 test: 0.2150922 best: 0.2150922 (6500) total: 6m 16s remaining: 20m 43s 6600: learn: 0.1331652 test: 0.2150781 best: 0.2150741 (6580) total: 6m 21s remaining: 20m 36s 6700: learn: 0.1322886 test: 0.2150534 best: 0.2150518 (6691) total: 6m 27s remaining: 20m 30s 6800: learn: 0.1313571 test: 0.2150445 best: 0.2150431 (6766) total: 6m 33s remaining: 20m 26s 6900: learn: 0.1304923 test: 0.2150271 best: 0.2150261 (6894) total: 6m 39s remaining: 20m 20s 7000: learn: 0.1296811 test: 0.2150225 best: 0.2150218 (6933) total: 6m 44s remaining: 20m 13s 7100: learn: 0.1288370 test: 0.2150057 best: 0.2150057 (7100) total: 6m 50s remaining: 20m 8s 7200: learn: 0.1279664 test: 0.2150026 best: 0.2149933 (7130) total: 6m 56s remaining: 20m 1s 7300: learn: 0.1271221 test: 0.2150052 best: 0.2149933 (7130) total: 7m 2s remaining: 19m 56s 7400: learn: 0.1262447 test: 0.2150003 best: 0.2149933 (7130) total: 7m 7s remaining: 19m 51s 7500: learn: 0.1254249 test: 0.2149963 best: 0.2149933 (7130) total: 7m 13s remaining: 19m 45s 7600: learn: 0.1245860 test: 0.2149678 best: 0.2149663 (7598) total: 7m 19s remaining: 19m 38s 7700: learn: 0.1237308 test: 0.2149614 best: 0.2149613 (7670) total: 7m 24s remaining: 19m 32s 7800: learn: 0.1229083 test: 0.2149524 best: 0.2149508 (7781) total: 7m 30s remaining: 19m 26s 7900: learn: 0.1220695 test: 0.2149601 best: 0.2149470 (7809) total: 7m 36s remaining: 19m 22s 8000: learn: 0.1212441 test: 0.2149425 best: 0.2149416 (7999) total: 7m 42s remaining: 19m 16s 8100: learn: 0.1204482 test: 0.2149245 best: 0.2149245 (8100) total: 7m 48s remaining: 19m 10s 8200: learn: 0.1196467 test: 0.2149143 best: 0.2149103 (8194) total: 7m 53s remaining: 19m 4s 8300: learn: 0.1188850 test: 0.2149177 best: 0.2149082 (8217) total: 7m 59s remaining: 18m 57s 8400: learn: 0.1181014 test: 0.2149274 best: 0.2149082 (8217) total: 8m 5s remaining: 18m 51s 8500: learn: 0.1173241 test: 0.2149441 best: 0.2149082 (8217) total: 8m 11s remaining: 18m 46s 8600: learn: 0.1165460 test: 0.2149354 best: 0.2149082 (8217) total: 8m 17s remaining: 18m 41s 8700: learn: 0.1158378 test: 0.2149264 best: 0.2149082 (8217) total: 8m 22s remaining: 18m 34s 8800: learn: 0.1151291 test: 0.2149128 best: 0.2149082 (8217) total: 8m 28s remaining: 18m 28s 8900: learn: 0.1143504 test: 0.2149062 best: 0.2149061 (8899) total: 8m 33s remaining: 18m 22s 9000: learn: 0.1135881 test: 0.2148948 best: 0.2148944 (8997) total: 8m 39s remaining: 18m 16s 9100: learn: 0.1128488 test: 0.2149008 best: 0.2148944 (8997) total: 8m 45s remaining: 18m 12s 9200: learn: 0.1121065 test: 0.2149064 best: 0.2148944 (8997) total: 8m 51s remaining: 18m 6s 9300: learn: 0.1113883 test: 0.2149018 best: 0.2148944 (8997) total: 8m 57s remaining: 17m 59s 9400: learn: 0.1106625 test: 0.2148942 best: 0.2148921 (9359) total: 9m 2s remaining: 17m 53s 9500: learn: 0.1099736 test: 0.2149004 best: 0.2148905 (9441) total: 9m 8s remaining: 17m 47s 9600: learn: 0.1092582 test: 0.2149066 best: 0.2148905 (9441) total: 9m 14s remaining: 17m 41s 9700: learn: 0.1085415 test: 0.2149110 best: 0.2148905 (9441) total: 9m 19s remaining: 17m 36s 9800: learn: 0.1078341 test: 0.2149177 best: 0.2148905 (9441) total: 9m 25s remaining: 17m 30s 9900: learn: 0.1071338 test: 0.2149166 best: 0.2148905 (9441) total: 9m 31s remaining: 17m 24s 10000: learn: 0.1065034 test: 0.2149291 best: 0.2148905 (9441) total: 9m 37s remaining: 17m 18s bestTest = 0.214890481 bestIteration = 9441 Shrink model to first 9442 iterations. Fold 3 | 09:55 | 9441 trees | Score = 0.79676 0: learn: 0.6791819 test: 0.6791714 best: 0.6791714 (0) total: 62ms remaining: 28m 54s 100: learn: 0.2716778 test: 0.2726303 best: 0.2726303 (100) total: 5.86s remaining: 26m 59s 200: learn: 0.2380137 test: 0.2405723 best: 0.2405723 (200) total: 12.2s remaining: 28m 300: learn: 0.2278132 test: 0.2320729 best: 0.2320729 (300) total: 18.1s remaining: 27m 43s 400: learn: 0.2223262 test: 0.2283289 best: 0.2283289 (400) total: 24s remaining: 27m 30s 500: learn: 0.2183863 test: 0.2260689 best: 0.2260689 (500) total: 29.5s remaining: 27m 600: learn: 0.2151291 test: 0.2244898 best: 0.2244898 (600) total: 35.5s remaining: 26m 56s 700: learn: 0.2122307 test: 0.2233024 best: 0.2233024 (700) total: 41.2s remaining: 26m 45s 800: learn: 0.2097304 test: 0.2223661 best: 0.2223661 (800) total: 47.5s remaining: 26m 54s 900: learn: 0.2073933 test: 0.2216285 best: 0.2216285 (900) total: 53.4s remaining: 26m 46s 1000: learn: 0.2052585 test: 0.2210236 best: 0.2210236 (1000) total: 58.8s remaining: 26m 26s 1100: learn: 0.2031697 test: 0.2205296 best: 0.2205296 (1100) total: 1m 4s remaining: 26m 19s 1200: learn: 0.2012590 test: 0.2200947 best: 0.2200947 (1200) total: 1m 10s remaining: 26m 3s 1300: learn: 0.1993650 test: 0.2197192 best: 0.2197192 (1300) total: 1m 15s remaining: 25m 56s 1400: learn: 0.1975051 test: 0.2194050 best: 0.2194050 (1400) total: 1m 21s remaining: 25m 42s 1500: learn: 0.1956982 test: 0.2191100 best: 0.2191100 (1500) total: 1m 27s remaining: 25m 36s 1600: learn: 0.1938946 test: 0.2188298 best: 0.2188292 (1597) total: 1m 33s remaining: 25m 38s 1700: learn: 0.1921411 test: 0.2186138 best: 0.2186138 (1700) total: 1m 39s remaining: 25m 33s 1800: learn: 0.1904881 test: 0.2184089 best: 0.2184089 (1800) total: 1m 44s remaining: 25m 22s 1900: learn: 0.1888548 test: 0.2182280 best: 0.2182280 (1900) total: 1m 50s remaining: 25m 15s 2000: learn: 0.1872993 test: 0.2180642 best: 0.2180642 (2000) total: 1m 55s remaining: 25m 4s 2100: learn: 0.1857231 test: 0.2179005 best: 0.2178989 (2097) total: 2m 2s remaining: 25m 4s 2200: learn: 0.1842326 test: 0.2177677 best: 0.2177677 (2200) total: 2m 7s remaining: 24m 57s 2300: learn: 0.1828134 test: 0.2176604 best: 0.2176604 (2300) total: 2m 13s remaining: 24m 51s 2400: learn: 0.1813027 test: 0.2175185 best: 0.2175179 (2399) total: 2m 18s remaining: 24m 41s 2500: learn: 0.1799160 test: 0.2174063 best: 0.2174063 (2500) total: 2m 24s remaining: 24m 36s 2600: learn: 0.1784561 test: 0.2173033 best: 0.2173033 (2600) total: 2m 30s remaining: 24m 28s 2700: learn: 0.1770173 test: 0.2172087 best: 0.2172087 (2700) total: 2m 37s remaining: 24m 31s 2800: learn: 0.1756181 test: 0.2171261 best: 0.2171261 (2800) total: 2m 42s remaining: 24m 25s 2900: learn: 0.1742843 test: 0.2170294 best: 0.2170282 (2895) total: 2m 48s remaining: 24m 16s 3000: learn: 0.1729695 test: 0.2169580 best: 0.2169580 (3000) total: 2m 54s remaining: 24m 10s 3100: learn: 0.1717441 test: 0.2168911 best: 0.2168911 (3100) total: 2m 59s remaining: 24m 1s 3200: learn: 0.1703941 test: 0.2168169 best: 0.2168155 (3199) total: 3m 5s remaining: 23m 56s 3300: learn: 0.1691239 test: 0.2167541 best: 0.2167541 (3300) total: 3m 11s remaining: 23m 53s 3400: learn: 0.1678480 test: 0.2167035 best: 0.2167035 (3400) total: 3m 17s remaining: 23m 47s 3500: learn: 0.1665836 test: 0.2166058 best: 0.2166057 (3498) total: 3m 22s remaining: 23m 39s 3600: learn: 0.1653981 test: 0.2165486 best: 0.2165486 (3599) total: 3m 28s remaining: 23m 32s 3700: learn: 0.1642256 test: 0.2165003 best: 0.2165003 (3700) total: 3m 33s remaining: 23m 24s 3800: learn: 0.1630228 test: 0.2164259 best: 0.2164259 (3800) total: 3m 39s remaining: 23m 18s 3900: learn: 0.1618181 test: 0.2163589 best: 0.2163589 (3900) total: 3m 45s remaining: 23m 15s 4000: learn: 0.1606510 test: 0.2163154 best: 0.2163154 (4000) total: 3m 51s remaining: 23m 9s 4100: learn: 0.1595218 test: 0.2162808 best: 0.2162800 (4096) total: 3m 57s remaining: 23m 2s 4200: learn: 0.1583397 test: 0.2162371 best: 0.2162371 (4200) total: 4m 3s remaining: 22m 57s 4300: learn: 0.1571928 test: 0.2161981 best: 0.2161981 (4300) total: 4m 8s remaining: 22m 49s 4400: learn: 0.1560129 test: 0.2161546 best: 0.2161498 (4394) total: 4m 14s remaining: 22m 43s 4500: learn: 0.1548678 test: 0.2161095 best: 0.2161095 (4499) total: 4m 20s remaining: 22m 39s 4600: learn: 0.1537756 test: 0.2160674 best: 0.2160663 (4595) total: 4m 26s remaining: 22m 33s 4700: learn: 0.1526616 test: 0.2160430 best: 0.2160411 (4695) total: 4m 31s remaining: 22m 26s 4800: learn: 0.1515711 test: 0.2159967 best: 0.2159963 (4798) total: 4m 37s remaining: 22m 20s 4900: learn: 0.1504769 test: 0.2159731 best: 0.2159731 (4900) total: 4m 43s remaining: 22m 14s 5000: learn: 0.1493968 test: 0.2159667 best: 0.2159626 (4990) total: 4m 48s remaining: 22m 7s 5100: learn: 0.1483408 test: 0.2159246 best: 0.2159218 (5090) total: 4m 55s remaining: 22m 5s 5200: learn: 0.1472730 test: 0.2159049 best: 0.2158988 (5175) total: 5m remaining: 21m 58s 5300: learn: 0.1462379 test: 0.2158672 best: 0.2158672 (5300) total: 5m 6s remaining: 21m 52s 5400: learn: 0.1452490 test: 0.2158456 best: 0.2158456 (5400) total: 5m 11s remaining: 21m 45s 5500: learn: 0.1442221 test: 0.2158317 best: 0.2158300 (5489) total: 5m 17s remaining: 21m 39s 5600: learn: 0.1432222 test: 0.2158154 best: 0.2158151 (5595) total: 5m 23s remaining: 21m 33s 5700: learn: 0.1422560 test: 0.2157759 best: 0.2157740 (5687) total: 5m 29s remaining: 21m 29s 5800: learn: 0.1412521 test: 0.2157759 best: 0.2157696 (5715) total: 5m 35s remaining: 21m 22s 5900: learn: 0.1402275 test: 0.2157585 best: 0.2157577 (5899) total: 5m 40s remaining: 21m 16s 6000: learn: 0.1392428 test: 0.2157357 best: 0.2157354 (5998) total: 5m 46s remaining: 21m 10s 6100: learn: 0.1382889 test: 0.2156968 best: 0.2156961 (6085) total: 5m 52s remaining: 21m 4s 6200: learn: 0.1372936 test: 0.2156935 best: 0.2156791 (6136) total: 5m 58s remaining: 21m 6300: learn: 0.1363213 test: 0.2156703 best: 0.2156703 (6300) total: 6m 4s remaining: 20m 54s 6400: learn: 0.1353630 test: 0.2156661 best: 0.2156637 (6398) total: 6m 9s remaining: 20m 48s 6500: learn: 0.1344365 test: 0.2156632 best: 0.2156614 (6481) total: 6m 15s remaining: 20m 42s 6600: learn: 0.1334934 test: 0.2156381 best: 0.2156371 (6599) total: 6m 21s remaining: 20m 35s 6700: learn: 0.1326188 test: 0.2156194 best: 0.2156150 (6674) total: 6m 27s remaining: 20m 30s 6800: learn: 0.1316889 test: 0.2156070 best: 0.2156048 (6774) total: 6m 33s remaining: 20m 26s 6900: learn: 0.1307244 test: 0.2155944 best: 0.2155944 (6900) total: 6m 38s remaining: 20m 19s 7000: learn: 0.1298473 test: 0.2155744 best: 0.2155681 (6975) total: 6m 44s remaining: 20m 13s 7100: learn: 0.1290104 test: 0.2155576 best: 0.2155556 (7092) total: 6m 50s remaining: 20m 6s 7200: learn: 0.1281102 test: 0.2155483 best: 0.2155483 (7200) total: 6m 55s remaining: 20m 1s 7300: learn: 0.1272382 test: 0.2155643 best: 0.2155447 (7209) total: 7m 1s remaining: 19m 54s 7400: learn: 0.1263787 test: 0.2155680 best: 0.2155447 (7209) total: 7m 7s remaining: 19m 50s 7500: learn: 0.1255063 test: 0.2155634 best: 0.2155447 (7209) total: 7m 13s remaining: 19m 44s 7600: learn: 0.1246314 test: 0.2155831 best: 0.2155447 (7209) total: 7m 19s remaining: 19m 38s 7700: learn: 0.1237732 test: 0.2155733 best: 0.2155447 (7209) total: 7m 24s remaining: 19m 31s 7800: learn: 0.1229562 test: 0.2155820 best: 0.2155447 (7209) total: 7m 30s remaining: 19m 26s bestTest = 0.2155447359 bestIteration = 7209 Shrink model to first 7210 iterations. Fold 4 | 07:45 | 7209 trees | Score = 0.79327 OOF Score: 0.79338
0
model.get_all_params()
{'nan_mode': 'Min',
'gpu_ram_part': 0.95,
'eval_metric': 'Logloss',
'iterations': 28000,
'leaf_estimation_method': 'Newton',
'observations_to_bootstrap': 'TestOnly',
'od_pval': 0,
'grow_policy': 'SymmetricTree',
'boosting_type': 'Plain',
'feature_border_type': 'GreedyLogSum',
'bayesian_matrix_reg': 0.10000000149011612,
'devices': '-1',
'pinned_memory_bytes': '104857600',
'force_unit_auto_pair_weights': False,
'l2_leaf_reg': 3,
'random_strength': 1,
'od_type': 'Iter',
'rsm': 1,
'boost_from_average': False,
'gpu_cat_features_storage': 'GpuRam',
'fold_size_loss_normalization': False,
'model_size_reg': 0.5,
'pool_metainfo_options': {'tags': {}},
'use_best_model': True,
'meta_l2_frequency': 0,
'od_wait': 600,
'class_names': [0, 1],
'random_seed': 0,
'depth': 10,
'border_count': 128,
'min_fold_size': 100,
'data_partition': 'DocParallel',
'bagging_temperature': 1,
'classes_count': 0,
'auto_class_weights': 'None',
'leaf_estimation_backtracking': 'AnyImprovement',
'best_model_min_trees': 1,
'min_data_in_leaf': 1,
'add_ridge_penalty_to_loss_function': False,
'loss_function': 'Logloss',
'learning_rate': 0.009999999776482582,
'meta_l2_exponent': 1,
'score_function': 'Cosine',
'task_type': 'GPU',
'leaf_estimation_iterations': 10,
'bootstrap_type': 'Bayesian',
'max_leaves': 1024}
top_n = plot_feature_importance(model.feature_importances_, features,'CatBoost', n=80, figsize=(20,12))
print(f'Mean CV score for {MAX_DEPTH} max_depth: {np.mean(score_list)}')
Mean CV score for 10 max_depth: 0.7933820749798006
# 5개 데이터 셋에 대한 평
print(f'LGBM baseline: {np.mean([0.79605, 0.79674, 0.79362, 0.79283, 0.79442])}')
LGBM baseline: 0.794732
np.exp((np.log(y_pred_list[0])+np.log(y_pred_list[1])+np.log(y_pred_list[2])+np.log(y_pred_list[3])+np.log(y_pred_list[4]))/5)
array([0.01859136, 0.00176784, 0.03999009, ..., 0.3929696 , 0.4208918 ,
0.04277044])
(y_pred_list[0]+y_pred_list[1]+y_pred_list[2]+y_pred_list[3]+y_pred_list[4])/5
array([0.01866999, 0.00177495, 0.04032808, ..., 0.39607875, 0.42188451,
0.04302383])
np.mean(y_pred_list[0])
0.25215407968337267
list(top_n)
['P_2_last', 'B_1_last', 'D_39_last', 'B_4_last', 'R_1_avg', 'S_3_avg', 'D_48_last', 'R_1_last', 'D_44_last', 'B_9_last', 'B_3_last', 'D_46_last', 'S_3_last', 'D_51_last', 'D_43_avg', 'B_2_last', 'D_39_max', 'B_11_last', 'B_5_last', 'R_2_last', 'D_42_max', 'D_43_last', 'R_3_avg', 'P_2_min', 'D_41_last', 'S_15_avg', 'B_4_max', 'D_49_last', 'D_45_max', 'R_1_max', 'D_42_avg', 'P_3_min', 'P_3_last', 'D_50_last', 'D_47_last', 'P_3_avg', 'D_47_max', 'S_23_last', 'S_3_max', 'D_46_min', 'B_9_avg', 'B_7_last', 'P_2_avg', 'B_37_last', 'D_62_min', 'B_23_last', 'S_26_max', 'D_112_last', 'B_8_avg', 'P_2_max', 'B_9_max', 'D_47_avg', 'D_42_min', 'D_52_last', 'B_28_min', 'D_66_avg', 'R_27_min', 'B_18_last', 'B_17_last', 'S_12_avg', 'R_6_last', 'S_12_min', 'B_10_last', 'B_26_last', 'B_2_min', 'S_9_avg', 'D_52_max', 'B_40_max', 'D_50_avg', 'D_61_last', 'S_11_avg', 'D_45_avg', 'S_3_min', 'S_24_last', 'D_53_last', 'D_48_min', 'P_3_max', 'D_46_avg', 'S_12_max', 'S_27_max']
sub = pd.DataFrame({'customer_ID': test.index,
'prediction': np.exp((np.log(y_pred_list[0])+np.log(y_pred_list[1])+np.log(y_pred_list[2])+
np.log(y_pred_list[3])+np.log(y_pred_list[4]))/5)})
sub.to_csv('./submission.csv', index=False)
sub
| customer_ID | prediction | |
|---|---|---|
| 0 | 00000469ba478561f23a92a868bd366de6f6527a684c9a... | 0.018591 |
| 1 | 00001bf2e77ff879fab36aa4fac689b9ba411dae63ae39... | 0.001768 |
| 2 | 0000210045da4f81e5f122c6bde5c2a617d03eef67f82c... | 0.039990 |
| 3 | 00003b41e58ede33b8daf61ab56d9952f17c9ad1c3976c... | 0.221594 |
| 4 | 00004b22eaeeeb0ec976890c1d9bfc14fd9427e98c4ee9... | 0.828064 |
| ... | ... | ... |
| 924616 | ffff952c631f2c911b8a2a8ca56ea6e656309a83d2f64c... | 0.010519 |
| 924617 | ffffcf5df59e5e0bba2a5ac4578a34e2b5aa64a1546cd3... | 0.799683 |
| 924618 | ffffd61f098cc056dbd7d2a21380c4804bbfe60856f475... | 0.392970 |
| 924619 | ffffddef1fc3643ea179c93245b68dca0f36941cd83977... | 0.420892 |
| 924620 | fffffa7cf7e453e1acc6a1426475d5cb9400859f82ff61... | 0.042770 |
924621 rows × 2 columns