AI - H2O - 安装与运行
安装的要求
H2O的安装对操作系统、编程语言和浏览器有具体的要求。
详情请查看官方信息
下载H2O
示例 - 在CentOS7.5中直接运行
查看系统及Java信息
[Anliven@localhost ~]$ uname -a
Linux localhost.localdomain 3.10.0-957.el7.x86_64 #1 SMP Thu Nov 8 23:39:32 UTC 2018 x86_64 x86_64 x86_64 GNU/Linux
[Anliven@localhost ~]$
[Anliven@localhost ~]$ cat /etc/system-release
CentOS Linux release 7.5.1804 (Core)
[Anliven@localhost ~]$
[Anliven@localhost ~]$ java -version
openjdk version "1.8.0_161"
OpenJDK Runtime Environment (build 1.8.0_161-b14)
OpenJDK 64-Bit Server VM (build 25.161-b14, mixed mode)
[Anliven@localhost ~]$
运行H2O
通过java -jar h2o.jar -ip <IP_Address> -port <PortNumber>命令运行H2O
[Anliven@localhost h2o-3.24.0.5]$ pwd
/home/Anliven/Downloads/h2o-3.24.0.5
[Anliven@localhost h2o-3.24.0.5]$
[Anliven@localhost h2o-3.24.0.5]$ ll
total 127012
drwxr-xr-x 3 Anliven Anliven 18 Jun 19 08:19 bindings
-rw-r--r-- 1 Anliven Anliven 130056596 Jun 19 08:19 h2o.jar
drwxr-xr-x 2 Anliven Anliven 47 Jun 19 08:19 python
drwxr-xr-x 2 Anliven Anliven 33 Jun 19 08:19 R
[Anliven@localhost h2o-3.24.0.5]$
[Anliven@localhost h2o-3.24.0.5]$ java -jar h2o.jar -ip 192.168.16.101 -port 54321
06-21 23:44:41.564 192.168.16.101:54321 4039 main INFO: ----- H2O started -----
06-21 23:44:41.582 192.168.16.101:54321 4039 main INFO: Build git branch: rel-yates
06-21 23:44:41.582 192.168.16.101:54321 4039 main INFO: Build git hash: b9cd4d5bcd44a4949ca8c677c5e54c10ee72c968
06-21 23:44:41.583 192.168.16.101:54321 4039 main INFO: Build git describe: jenkins-3.24.0.4-66-gb9cd4d5
06-21 23:44:41.583 192.168.16.101:54321 4039 main INFO: Build project version: 3.24.0.5
06-21 23:44:41.583 192.168.16.101:54321 4039 main INFO: Build age: 2 days
06-21 23:44:41.583 192.168.16.101:54321 4039 main INFO: Built by: 'jenkins'
06-21 23:44:41.583 192.168.16.101:54321 4039 main INFO: Built on: '2019-06-18 23:52:14'
06-21 23:44:41.583 192.168.16.101:54321 4039 main INFO: Found H2O Core extensions: [Watchdog, XGBoost, KrbStandalone]
06-21 23:44:41.584 192.168.16.101:54321 4039 main INFO: Processed H2O arguments: [-ip, 192.168.16.101, -port, 54321]
06-21 23:44:41.584 192.168.16.101:54321 4039 main INFO: Java availableProcessors: 4
06-21 23:44:41.584 192.168.16.101:54321 4039 main INFO: Java heap totalMemory: 240.0 MB
06-21 23:44:41.584 192.168.16.101:54321 4039 main INFO: Java heap maxMemory: 3.45 GB
06-21 23:44:41.584 192.168.16.101:54321 4039 main INFO: Java version: Java 1.8.0_161 (from Oracle Corporation)
06-21 23:44:41.584 192.168.16.101:54321 4039 main INFO: JVM launch parameters: []
06-21 23:44:41.585 192.168.16.101:54321 4039 main INFO: OS version: Linux 3.10.0-957.el7.x86_64 (amd64)
06-21 23:44:41.585 192.168.16.101:54321 4039 main INFO: Machine physical memory: 15.51 GB
06-21 23:44:41.585 192.168.16.101:54321 4039 main INFO: Machine locale: en_US
06-21 23:44:41.585 192.168.16.101:54321 4039 main INFO: X-h2o-cluster-id: 1561131880800
06-21 23:44:41.585 192.168.16.101:54321 4039 main INFO: User name: 'Anliven'
06-21 23:44:41.585 192.168.16.101:54321 4039 main INFO: IPv6 stack selected: false
06-21 23:44:41.585 192.168.16.101:54321 4039 main INFO: Network interface is down: name:virbr0 (virbr0)
06-21 23:44:41.585 192.168.16.101:54321 4039 main INFO: Possible IP Address: enp0s8 (enp0s8), fe80:0:0:0:cfdd:6281:f738:fba%enp0s8
06-21 23:44:41.586 192.168.16.101:54321 4039 main INFO: Possible IP Address: enp0s8 (enp0s8), 192.168.16.101
06-21 23:44:41.586 192.168.16.101:54321 4039 main INFO: Possible IP Address: enp0s3 (enp0s3), fe80:0:0:0:c48f:c289:276:2308%enp0s3
06-21 23:44:41.586 192.168.16.101:54321 4039 main INFO: Possible IP Address: enp0s3 (enp0s3), 10.0.2.15
06-21 23:44:41.586 192.168.16.101:54321 4039 main INFO: Possible IP Address: lo (lo), 0:0:0:0:0:0:0:1%lo
06-21 23:44:41.586 192.168.16.101:54321 4039 main INFO: Possible IP Address: lo (lo), 127.0.0.1
06-21 23:44:41.586 192.168.16.101:54321 4039 main INFO: H2O node running in unencrypted mode.
06-21 23:44:41.588 192.168.16.101:54321 4039 main INFO: Internal communication uses port: 54322
06-21 23:44:41.588 192.168.16.101:54321 4039 main INFO: Listening for HTTP and REST traffic on http://192.168.16.101:54321/
06-21 23:44:41.589 192.168.16.101:54321 4039 main INFO: H2O cloud name: 'Anliven' on /192.168.16.101:54321, static configuration based on -flatfile null
06-21 23:44:41.589 192.168.16.101:54321 4039 main INFO: If you have trouble connecting, try SSH tunneling from your local machine (e.g., via port 55555):
06-21 23:44:41.589 192.168.16.101:54321 4039 main INFO: 1. Open a terminal and run 'ssh -L 55555:localhost:54321 Anliven@192.168.16.101'
06-21 23:44:41.589 192.168.16.101:54321 4039 main INFO: 2. Point your browser to http://localhost:55555
06-21 23:44:42.307 192.168.16.101:54321 4039 main INFO: Log dir: '/tmp/h2o-Anliven/h2ologs'
06-21 23:44:42.307 192.168.16.101:54321 4039 main INFO: Cur dir: '/home/Anliven/Downloads/h2o-3.24.0.5'
06-21 23:44:42.321 192.168.16.101:54321 4039 main INFO: Subsystem for distributed import from HTTP/HTTPS successfully initialized
06-21 23:44:42.322 192.168.16.101:54321 4039 main INFO: HDFS subsystem successfully initialized
06-21 23:44:42.327 192.168.16.101:54321 4039 main INFO: S3 subsystem successfully initialized
06-21 23:44:42.352 192.168.16.101:54321 4039 main INFO: GCS subsystem successfully initialized
06-21 23:44:42.352 192.168.16.101:54321 4039 main INFO: Flow dir: '/home/Anliven/h2oflows'
06-21 23:44:42.372 192.168.16.101:54321 4039 main INFO: Cloud of size 1 formed [/192.168.16.101:54321]
06-21 23:44:42.386 192.168.16.101:54321 4039 main INFO: Registered parsers: [GUESS, ARFF, XLS, SVMLight, AVRO, PARQUET, CSV]
06-21 23:44:42.387 192.168.16.101:54321 4039 main INFO: Watchdog extension initialized
06-21 23:44:42.387 192.168.16.101:54321 4039 main INFO: XGBoost extension initialized
06-21 23:44:42.388 192.168.16.101:54321 4039 main INFO: KrbStandalone extension initialized
06-21 23:44:42.388 192.168.16.101:54321 4039 main INFO: Registered 3 core extensions in: 327ms
06-21 23:44:42.389 192.168.16.101:54321 4039 main INFO: Registered H2O core extensions: [Watchdog, XGBoost, KrbStandalone]
06-21 23:44:42.625 192.168.16.101:54321 4039 main INFO: Found XGBoost backend with library: xgboost4j_gpu
06-21 23:44:42.625 192.168.16.101:54321 4039 main INFO: XGBoost supported backends: [WITH_GPU, WITH_OMP]
06-21 23:44:42.788 192.168.16.101:54321 4039 main INFO: Registered: 174 REST APIs in: 399ms
06-21 23:44:42.788 192.168.16.101:54321 4039 main INFO: Registered REST API extensions: [Amazon S3, XGBoost, Algos, AutoML, Core V3, Core V4]
06-21 23:44:43.005 192.168.16.101:54321 4039 main INFO: Registered: 249 schemas in 216ms
06-21 23:44:43.005 192.168.16.101:54321 4039 main INFO: H2O started in 2195ms
06-21 23:44:43.005 192.168.16.101:54321 4039 main INFO:
06-21 23:44:43.005 192.168.16.101:54321 4039 main INFO: Open H2O Flow in your web browser: http://192.168.16.101:54321
06-21 23:44:43.006 192.168.16.101:54321 4039 main INFO:
H2O Flow的web页面:
h2o.jar的帮助信息
执行java -jar h2o.jar --help命令显示帮助信息。
问题处理
问题1：无法打开H2O的web页面
处理方法：
查看启动日志,如果看到类似"http://192.168.16.101:54321"信息说明H2O已经成功启动,
那么此时问题的原因应该与网络相关,需要检查防火墙/代理/路由等相关网络设置.
建议先检查防火墙的设置。可以关闭防火墙并设置为开机不启动，也可以将H2O的web服务加入到防火墙的规则中。
[root@localhost ~]# firewall-cmd --state
running
[root@localhost ~]# systemctl stop firewalld && systemctl disable firewalld
Removed symlink /etc/systemd/system/multi-user.target.wants/firewalld.service.
Removed symlink /etc/systemd/system/dbus-org.fedoraproject.FirewallD1.service.
[root@localhost ~]#
问题2：H2O的启动日志中显示“Failed to determine IP, falling back to localhost”信息
执行java -jar h2o.jar后，H2O的启动日志中显示有“Failed to determine IP, falling back to localhost”信息
处理方法：通过java -jar h2o.jar -ip <IP_Address> -port <PortNumber>命令指定IP地址和端口来运行H2O。
示例 - 在Anaconda3环境中安装H2O并运行
官网信息
使用的命令
conda create -n h2o pip python=3.6 # 创建Python3.6的虚拟环境
conda activate h2o # 激活并进入虚拟环境
pip install -U h2o # 在虚拟环境中安装h2o，参数-U表明要升级安装任何依赖项
安装完成后的包列表
(h2o) C:\Users\guowli>pip list
Package Version
------------ --------
certifi 2019.3.9
chardet 3.0.4
colorama 0.4.1
future 0.17.1
h2o 3.24.0.5
idna 2.8
pip 19.1.1
requests 2.22.0
setuptools 41.0.1
tabulate 0.8.3
urllib3 1.25.3
wheel 0.33.4
wincertstore 0.2
(h2o) C:\Users\guowli>
运行初始化
(h2o) C:\Users\guowli>python
Python 3.6.8 |Anaconda, Inc.| (default, Feb 21 2019, 18:30:04) [MSC v.1916 64 bit (AMD64)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> import h2o # 导入模块
>>> h2o.init() # 将显示启动的相关信息,表格中包括可用的节点个数/存储空间/内核个数等信息
Checking whether there is an H2O instance running at http://localhost:54321 ..... not found.
Attempting to start a local H2O server...
; Java HotSpot(TM) Client VM (build 25.152-b16, mixed mode)
C:\Office-Tools\Anaconda3\envs\h2o\lib\site-packages\h2o\backend\server.py:369: UserWarning: You have a 32-bit version of Java. H2O works best with 64-bit Java.
Please download the latest 64-bit Java SE JDK from Oracle.
warn(" You have a 32-bit version of Java. H2O works best with 64-bit Java.\n"
Starting server from C:\Office-Tools\Anaconda3\envs\h2o\lib\site-packages\h2o\backend\bin\h2o.jar
Ice root: C:\Users\guowli\AppData\Local\Temp\tmpydo64nu9
JVM stdout: C:\Users\guowli\AppData\Local\Temp\tmpydo64nu9\h2o_guowli_started_from_python.out
JVM stderr: C:\Users\guowli\AppData\Local\Temp\tmpydo64nu9\h2o_guowli_started_from_python.err
Server is running at http://127.0.0.1:54321
Connecting to H2O server at http://127.0.0.1:54321 ... successful.
-------------------------- ------------------------------------------
H2O cluster uptime: 01 secs
H2O cluster timezone: Asia/Shanghai
H2O data parsing timezone: UTC
H2O cluster version: 3.24.0.5 # H2O版本
H2O cluster version age: 10 hours and 43 minutes
H2O cluster name: H2O_from_python_guowli_76mkk5
H2O cluster total nodes: 1 # 集群节点个数
H2O cluster free memory: 247.5 Mb
H2O cluster total cores: 8 # 集群内核个数
H2O cluster allowed cores: 8 # 集群可用内核个数
H2O cluster status: accepting new members, healthy
H2O connection url: http://127.0.0.1:54321 # Web地址
H2O connection proxy:
H2O internal security: False
H2O API Extensions: Amazon S3, Algos, AutoML, Core V3, Core V4
Python version: 3.6.8 final # Python版本
-------------------------- ------------------------------------------
>>>
注意:
- 默认情况下,H2O实例允许使用所有内核, 并且通常需要25%的系统存储空间.
- 可以通过类似
h2o.init(nthreads=2,max_mem_size=4)命令指定相关启动配置. - 通过
h2o.shutdown()1命令关闭H2O实例.
运行官方Demo
>>> h2o.demo("glm")
-------------------------------------------------------------------------------
Demo of H2O's Generalized Linear Estimator.
This demo uploads a dataset to h2o, parses it, and shows a description.
Then it divides the dataset into training and test sets, builds a GLM
from the training set, and makes predictions for the test set.
Finally, default performance metrics are displayed.
-------------------------------------------------------------------------------
>>> # Connect to H2O
>>> h2o.init()
Checking whether there is an H2O instance running at http://localhost:54321 . connected.
-------------------------- ------------------------------------------
H2O cluster uptime: 44 secs
H2O cluster timezone: Asia/Shanghai
H2O data parsing timezone: UTC
H2O cluster version: 3.24.0.5
H2O cluster version age: 10 hours and 44 minutes
H2O cluster name: H2O_from_python_guowli_76mkk5
H2O cluster total nodes: 1
H2O cluster free memory: 240.7 Mb
H2O cluster total cores: 8
H2O cluster allowed cores: 8
H2O cluster status: locked, healthy
H2O connection url: http://localhost:54321
H2O connection proxy:
H2O internal security: False
H2O API Extensions: Amazon S3, Algos, AutoML, Core V3, Core V4
Python version: 3.6.8 final
-------------------------- ------------------------------------------
>>> # Upload the prostate dataset that comes included in the h2o python package
>>> prostate = h2o.load_dataset("prostate")
Parse progress: |█████████████████████████████████████████████████████████| 100%
>>> # Print a description of the prostate data
>>> prostate.describe()
Rows:380
Cols:9
ID CAPSULE AGE RACE DPROS DCAPS PSA VOL GLEASON
------- ------------------ ------------------ ----------------- ------------------ ------------------ ------------------ ------------------ ------------------ ------------------
type int int int int int int real real int
mins 1.0 0.0 43.0 0.0 1.0 1.0 0.3 0.0 0.0
mean 190.5 0.4026315789473684 66.03947368421049 1.0868421052631572 2.2710526315789488 1.1078947368421048 15.408631578947375 15.812921052631573 6.3842105263157904
maxs 380.0 1.0 79.0 2.0 4.0 2.0 139.7 97.6 9.0
sigma 109.84079387914127 0.4910743389630552 6.527071269173311 0.3087732580252793 1.0001076181502861 0.3106564493514939 19.99757266856046 18.347619967271175 1.0919533744261092
zeros 0 227 0 3 0 0 0 167 2
missing 0 0 0 0 0 0 0 0 0
0 1.0 0.0 65.0 1.0 2.0 1.0 1.4 0.0 6.0
1 2.0 0.0 72.0 1.0 3.0 2.0 6.7 0.0 7.0
2 3.0 0.0 70.0 1.0 1.0 2.0 4.9 0.0 6.0
3 4.0 0.0 76.0 2.0 2.0 1.0 51.2 20.0 7.0
4 5.0 0.0 69.0 1.0 1.0 1.0 12.3 55.9 6.0
5 6.0 1.0 71.0 1.0 3.0 2.0 3.3 0.0 8.0
6 7.0 0.0 68.0 2.0 4.0 2.0 31.9 0.0 7.0
7 8.0 0.0 61.0 2.0 4.0 2.0 66.7 27.2 7.0
8 9.0 0.0 69.0 1.0 1.0 1.0 3.9 24.0 7.0
9 10.0 0.0 68.0 2.0 1.0 2.0 13.0 0.0 6.0
>>> # Randomly split the dataset into ~70/30, training/test sets
>>> train, test = prostate.split_frame(ratios=[0.70])
>>> # Convert the response columns to factors (for binary classification problems)
>>> train["CAPSULE"] = train["CAPSULE"].asfactor()
>>> test["CAPSULE"] = test["CAPSULE"].asfactor()
>>> # Build a (classification) GLM
>>> from h2o.estimators import H2OGeneralizedLinearEstimator
>>> prostate_glm = H2OGeneralizedLinearEstimator(family="binomial", alpha=[0.5])
>>> prostate_glm.train(x=["AGE", "RACE", "PSA", "VOL", "GLEASON"],
... y="CAPSULE", training_frame=train)
glm Model Build progress: |███████████████████████████████████████████████| 100%
>>> # Show the model
>>> prostate_glm.show()
Model Details
=============
H2OGeneralizedLinearEstimator : Generalized Linear Modeling
Model Key: GLM_model_python_1560911750112_1
ModelMetricsBinomialGLM: glm
** Reported on train data. **
MSE: 0.16734436667135488
RMSE: 0.40907745803375045
LogLoss: 0.5023661857779066
Null degrees of freedom: 271
Residual degrees of freedom: 266
Null deviance: 368.556956020097
Residual deviance: 273.28720506318115
AIC: 285.28720506318115
AUC: 0.8176339285714287
pr_auc: 0.7776373382337975
Gini: 0.6352678571428574
Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.30277329744729137:
0 1 Error Rate
----- --- --- ------- ------------
0 111 49 0.3063 (49.0/160.0)
1 20 92 0.1786 (20.0/112.0)
Total 131 141 0.2537 (69.0/272.0)
Maximum Metrics: Maximum metrics at their respective thresholds
metric threshold value idx
--------------------------- ----------- -------- -----
max f1 0.302773 0.727273 140
max f2 0.167286 0.807175 220
max f0point5 0.599644 0.742574 72
max accuracy 0.527291 0.768382 98
max precision 0.980771 1 0
max recall 0.0656329 1 252
max specificity 0.980771 1 0
max absolute_mcc 0.524337 0.516584 100
max min_per_class_accuracy 0.443324 0.741071 123
max mean_per_class_accuracy 0.302773 0.757589 140
Gains/Lift Table: Avg response rate: 41.18 %, avg score: 41.18 %
group cumulative_data_fraction lower_threshold lift cumulative_lift response_rate score cumulative_response_rate cumulative_score capture_rate cumulative_capture_rate gain cumulative_gain
-- ------- -------------------------- ----------------- -------- ----------------- --------------- --------- -------------------------- ------------------ -------------- ------------------------- -------- -----------------
1 0.0110294 0.975592 2.42857 2.42857 1 0.979539 1 0.979539 0.0267857 0.0267857 142.857 142.857
2 0.0220588 0.966995 2.42857 2.42857 1 0.971859 1 0.975699 0.0267857 0.0535714 142.857 142.857
3 0.0330882 0.961389 2.42857 2.42857 1 0.964036 1 0.971811 0.0267857 0.0803571 142.857 142.857
4 0.0404412 0.949559 2.42857 2.42857 1 0.956522 1 0.969032 0.0178571 0.0982143 142.857 142.857
5 0.0514706 0.922488 2.42857 2.42857 1 0.938832 1 0.96256 0.0267857 0.125 142.857 142.857
6 0.102941 0.863277 2.2551 2.34184 0.928571 0.889015 0.964286 0.925788 0.116071 0.241071 125.51 134.184
7 0.150735 0.709532 1.49451 2.07317 0.615385 0.790488 0.853659 0.882888 0.0714286 0.3125 49.4505 107.317
8 0.202206 0.634824 1.73469 1.98701 0.714286 0.665299 0.818182 0.827502 0.0892857 0.401786 73.4694 98.7013
9 0.301471 0.584551 1.5291 1.83624 0.62963 0.606812 0.756098 0.754835 0.151786 0.553571 52.9101 83.6237
10 0.400735 0.495188 1.25926 1.69332 0.518519 0.537514 0.697248 0.701003 0.125 0.678571 25.9259 69.3316
11 0.5 0.338356 1.07937 1.57143 0.444444 0.433575 0.647059 0.647911 0.107143 0.785714 7.93651 57.1429
12 0.599265 0.250821 0.719577 1.43032 0.296296 0.2807 0.588957 0.587085 0.0714286 0.857143 -28.0423 43.0324
13 0.698529 0.214874 0.269841 1.26541 0.111111 0.235682 0.521053 0.537149 0.0267857 0.883929 -73.0159 26.5414
14 0.797794 0.174605 0.62963 1.18631 0.259259 0.196557 0.488479 0.494771 0.0625 0.946429 -37.037 18.6307
15 0.897059 0.076389 0.359788 1.09485 0.148148 0.115647 0.45082 0.452819 0.0357143 0.982143 -64.0212 9.48478
16 1 0.000108149 0.173469 1 0.0714286 0.0540133 0.411765 0.411765 0.0178571 1 -82.6531 0
Scoring History:
timestamp duration iterations negative_log_likelihood objective
-- ------------------- ---------- ------------ ------------------------- -----------
2019-06-19 10:37:22 0.000 sec 0 184.278 0.677494
2019-06-19 10:37:22 0.012 sec 1 140.926 0.518611
2019-06-19 10:37:22 0.021 sec 2 136.838 0.503852
2019-06-19 10:37:22 0.022 sec 3 136.645 0.503224
2019-06-19 10:37:22 0.023 sec 4 136.644 0.503222
>>> # Predict on the test set and show the first ten predictions
>>> predictions = prostate_glm.predict(test)
>>> predictions.show()
glm prediction progress: |████████████████████████████████████████████████| 100%
predict p0 p1
--------- -------- ---------
1 0.457574 0.542426
1 0.189866 0.810134
1 0.419438 0.580562
1 0.521769 0.478231
1 0.375439 0.624561
0 0.927869 0.0721311
0 0.960693 0.0393066
0 0.700254 0.299746
0 0.714227 0.285773
0 0.778058 0.221942
[108 rows x 3 columns]
>>> # Show default performance metrics
>>> performance = prostate_glm.model_performance(test)
>>> performance.show()
ModelMetricsBinomialGLM: glm
** Reported on test data. **
MSE: 0.20621247932950715
RMSE: 0.45410624233708474
LogLoss: 0.5944711796848934
Null degrees of freedom: 107
Residual degrees of freedom: 102
Null deviance: 143.86304763240474
Residual deviance: 128.40577481193696
AIC: 140.40577481193696
AUC: 0.740444120859119
pr_auc: 0.6109686413835654
Gini: 0.4808882417182381
Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.2449261037363724:
0 1 Error Rate
----- --- --- ------- ------------
0 38 29 0.4328 (29.0/67.0)
1 6 35 0.1463 (6.0/41.0)
Total 44 64 0.3241 (35.0/108.0)
Maximum Metrics: Maximum metrics at their respective thresholds
metric threshold value idx
--------------------------- ----------- -------- -----
max f1 0.244926 0.666667 63
max f2 0.132351 0.795918 80
max f0point5 0.285773 0.59387 54
max accuracy 0.594262 0.694444 23
max precision 0.996946 1 0
max recall 0.0644647 1 98
max specificity 0.996946 1 0
max absolute_mcc 0.244926 0.415635 63
max min_per_class_accuracy 0.325993 0.682927 48
max mean_per_class_accuracy 0.244926 0.710411 63
Gains/Lift Table: Avg response rate: 37.96 %, avg score: 37.52 %
group cumulative_data_fraction lower_threshold lift cumulative_lift response_rate score cumulative_response_rate cumulative_score capture_rate cumulative_capture_rate gain cumulative_gain
-- ------- -------------------------- ----------------- -------- ----------------- --------------- --------- -------------------------- ------------------ -------------- ------------------------- -------- -----------------
1 0.0185185 0.979652 2.63415 2.63415 1 0.98867 1 0.98867 0.0487805 0.0487805 163.415 163.415
2 0.0277778 0.968501 2.63415 2.63415 1 0.969789 1 0.982377 0.0243902 0.0731707 163.415 163.415
3 0.037037 0.959393 2.63415 2.63415 1 0.960585 1 0.976929 0.0243902 0.097561 163.415 163.415
4 0.0462963 0.954581 2.63415 2.63415 1 0.954909 1 0.972525 0.0243902 0.121951 163.415 163.415
5 0.0555556 0.949862 2.63415 2.63415 1 0.953739 1 0.969394 0.0243902 0.146341 163.415 163.415
6 0.101852 0.799582 1.05366 1.91574 0.4 0.850374 0.727273 0.915294 0.0487805 0.195122 5.36585 91.5743
7 0.157407 0.658583 0.878049 1.5495 0.333333 0.710796 0.588235 0.843118 0.0487805 0.243902 -12.1951 54.9498
8 0.203704 0.598989 2.10732 1.67627 0.8 0.624421 0.636364 0.793414 0.097561 0.341463 110.732 67.6275
9 0.305556 0.538529 0.957871 1.43681 0.363636 0.562347 0.545455 0.716392 0.097561 0.439024 -4.21286 43.6807
10 0.398148 0.458199 1.31707 1.40896 0.5 0.510598 0.534884 0.668533 0.121951 0.560976 31.7073 40.8962
11 0.5 0.286173 1.67627 1.46341 0.636364 0.350458 0.555556 0.60374 0.170732 0.731707 67.6275 46.3415
12 0.601852 0.235707 1.19734 1.41839 0.454545 0.262231 0.538462 0.545946 0.121951 0.853659 19.7339 41.8386
13 0.694444 0.183515 0.526829 1.29951 0.2 0.212878 0.493333 0.501537 0.0487805 0.902439 -47.3171 29.9512
14 0.796296 0.095848 0.478936 1.19455 0.181818 0.143011 0.453488 0.455679 0.0487805 0.95122 -52.1064 19.4555
15 0.898148 0.0664361 0.239468 1.08625 0.0909091 0.0762976 0.412371 0.412656 0.0243902 0.97561 -76.0532 8.62459
16 1 0.000121128 0.239468 1 0.0909091 0.044715 0.37963 0.375181 0.0243902 1 -76.0532 0
---- End of Demo ----
>>>
AI - H2O - 安装与运行的更多相关文章
- AI - H2O - 第一个示例
1 - Iris数据集 Iris数据集是常用的机器学习分类实验数据集,特点是数据量很小,可以快速学习. 数据集包含150个数据集,分为3类,每类50个数据,每个数据包含4个属性. Sepal.Leng ...
- Python 2/3 安装与运行环境设置
Python 2/3 安装与运行环境设置: 1.Python 软件源:https://www.python.org/ 下载Win版本 https://www.python.org/downloa ...
- 使用工具安装，运行，停止，卸载Window服务
WSWinForm.exe介绍 WSWinForm.exe是我自己开发的一个实用的小工具,用于将任何EXE程序作为Windows服务运行.也就是说WSWinForm只是其注册程序的服务外壳,这个特性对 ...
- VMware10.06精简版安装后台运行
VMware10.06精简版安装时会出现一个安装功能选择菜单,里面有一条后台运行必选功能,一般人会跳过条.当你打算在服务器上用vmware时,一定要安装后台运行服务,否则你无法换出正在运行的后台虚拟机 ...
- phpstudy 出现“请自行检查是否安装VC9运行库？？”问题 -- 缺少vc9库
phpStudy是一款PHP调试环境的程序集成包,该程序包集成最新的Apache+PHP+MySQL+phpMyAdmin+ZendOptimizer,一次性安装,无须配置即可使用,是非常方便.好用的 ...
- JavaWeb学习之tomcat安装与运行、tomcat的目录结构、配置tomcat的管理用户、web项目目录、虚拟目录、虚拟主机（1）
1.tomcat安装与运行双击tomcat目录下的bin/startup.bat,启动之后,输入http://localhost:8080,出现安装成功的提示,表示安装tomcat成功 2.tomca ...
- 解决安装vc2005运行库时提示Command line option syntax error.Type Command/?for Help
安装vc2005运行库时提示 这是因为它要自解压到用户的临时文件夹下,如果用户名中带中文,就会报错. 简单的解决方法是,手动解压之,再安装 当然,你也可以修改用户名或者再新建个用户.
- 解决不安装VC运行库（VC2005，VC2008），程序运行出错的方法
因为VS2005以后程序采用了manifest的生成方式,所以发布的时候要和运行库一起发布.但是我们平时开发和发布的时候如果都要客户安装运行库,那就不太方便了.你可以Microsoft下载:http: ...
- win7 64下暗黑世界V1.1 服务器端及客户端的安装及运行 成功
原地址:http://bbs.gameres.com/thread_223717.html 屌丝一枚,没有MAC 更没有XCODE 潜心整了星期六与星期天两天才安装运行成功...只能说安装太复杂了,, ...
随机推荐
- C++ error C2015: too many characters in constant
错误原因:字符常量中的字符太多了. 错误分析: 单引号表示字符型常量. 一般的,单引号中必须有,也只能有一个字符(使用转义符时,转义符所表示的字符当作一个字符看待),如果单引号中的字符数多于4个,就会 ...
- 看加载的php.ini 和 phpinfo 配置路径
php -i | grep "phar.readonly"看当前值php -i | grep "php.ini" 看加载的php.ini是哪个
- 域渗透复盘(安洵CTF线下)
复盘线下域渗透环境Write Up 0x01 外网web到DMZ进域 外网web入口 joomla应用 192.168.0.5 反序列化打下来 GET /index.php HTTP/1.1 Ho ...
- Java 中Math常用方法
import java.text.SimpleDateFormat; import java.util.Date; public class Test4 { public static void ma ...
- ZR#989
ZR#989 先吐槽一下这个ZZ出题人,卡哈希表. 我就不写那个能过的类高精了,直接写哈希的题解 解法: 判断两个数相加结果是否等于第三个数, 可以直接用 hash判断． #include<io ...
- bootstrap table插件动态加载表头
这篇文章主要为大家详细介绍了bootstrap table插件动态加载表头,具有一定的参考价值,感兴趣的小伙伴们可以参考一下 bootstrap的table属性已经很熟悉了,最近遇到一个问题,犹豫 ...
- mongo最大连接数查看
进入客户端 mongo 输入查看命令 db.serverStatus().connections
- java输出程序运行时间
做了一个MapReduce的小练习,想测试一下程序运行时间: 代码: long start = System.currentTimeMillis(); /*运行的程序主体*/ long end = S ...
- Assignment5: 使用Visual Studio 进行可编码的UI测试
一.实验目的: 使用Visual Studio 可编码的UI测试功能创作自动化测试 二.实验原理:黑盒测试 三.实验需求:win8 app.Visual Studio Ultimate 2012/20 ...
- C平衡二叉树(AVL)创建和删除
AVL是最先发明的自平衡二叉查找树算法.在AVL中任何节点的两个儿子子树的高度最大差别为一,所以它也被称为高度平衡树,n个结点的AVL树最大深度约1.44log2n.查找.插入和删除在平均和最坏情况下 ...