Pythonで機械学習 - Modeling #機械学習

予測モデルの生成

必要なライブラリをインポートします。

import pandas as pd
from sklearn.ensemble import RandomForestClassifier

データファイルを読み込みます。

train = pd.read_csv("train_prep.csv")
test = pd.read_csv("test_prep.csv")

train.head(1)

	Unnamed: 0	PassengerId	Survived	Pclass	Name	Sex	Age	SibSp	Parch	Ticket	Fare	Cabin	Embarked	SexInt	AgeFillNa	FareFillNa	EmbarkedInt	NumFamily	IsAlone
0	0	1	0	3	Braund, Mr. Owen Harris	male	22.0	1	0	A/5 21171	7.25	NaN	S	0	22.0	7.25	0	1	0

学習用データの説明変数（インプット）と目的変数（アウトプット）を用意します。

expvars = ["Pclass","SexInt","AgeFillNa","FareFillNa","IsAlone"] # 説明変数のリスト
X_train = train.copy()[expvars] # 説明変数
Y_train = train["Survived"]# 目的変数

学習データによって学習させ、ランダムフォレストの予測モデルを構築します。

clf = RandomForestClassifier()
clf.fit(X_train, Y_train)

RandomForestClassifier(bootstrap=True, class_weight=None, criterion='gini',
            max_depth=None, max_features='auto', max_leaf_nodes=None,
            min_impurity_split=1e-07, min_samples_leaf=1,
            min_samples_split=2, min_weight_fraction_leaf=0.0,
            n_estimators=10, n_jobs=1, oob_score=False, random_state=None,
            verbose=0, warm_start=False)

予測モデルの確認

構築したモデルでテストデータの乗客の生存を予測します。

X_train = test.copy()[expvars] # 説明変数

prediction = clf.predict(X_train) # 生死の判定
pred_proba = clf.predict_proba(X_train) # 生存確率の判定

乗客データと合わせて可視化してみます。

test["PredictedSurvival"] = prediction
test["SurvivalProba"] = pred_proba[:,1]

test[["PassengerId","Pclass","Name","Sex","Age","SibSp","Parch","Fare","Embarked","PredictedSurvival","SurvivalProba"]]

	PassengerId	Pclass	Name	Sex	Age	SibSp	Parch	Fare	Embarked	PredictedSurvival	SurvivalProba
0	892	3	Kelly, Mr. James	male	34.5	0	0	7.8292	Q	0	0.000000
1	893	3	Wilkes, Mrs. James (Ellen Needs)	female	47.0	1	0	7.0000	S	0	0.300000
2	894	2	Myles, Mr. Thomas Francis	male	62.0	0	0	9.6875	Q	1	0.700000
3	895	3	Wirz, Mr. Albert	male	27.0	0	0	8.6625	S	1	0.700000
4	896	3	Hirvonen, Mrs. Alexander (Helga E Lindqvist)	female	22.0	1	1	12.2875	S	1	0.700000
5	897	3	Svensson, Mr. Johan Cervin	male	14.0	0	0	9.2250	S	0	0.200000
6	898	3	Connolly, Miss. Kate	female	30.0	0	0	7.6292	Q	0	0.100000
7	899	2	Caldwell, Mr. Albert Francis	male	26.0	1	1	29.0000	S	0	0.000000
8	900	3	Abrahim, Mrs. Joseph (Sophie Halaut Easu)	female	18.0	0	0	7.2292	C	1	1.000000
9	901	3	Davies, Mr. John Samuel	male	21.0	2	0	24.1500	S	0	0.000000
10	902	3	Ilieff, Mr. Ylio	male	NaN	0	0	7.8958	S	0	0.200000
11	903	1	Jones, Mr. Charles Cresson	male	46.0	0	0	26.0000	S	0	0.350000
12	904	1	Snyder, Mrs. John Pillsbury (Nelle Stevenson)	female	23.0	1	0	82.2667	S	1	1.000000
13	905	2	Howard, Mr. Benjamin	male	63.0	1	0	26.0000	S	0	0.000000
14	906	1	Chaffee, Mrs. Herbert Fuller (Carrie Constance...	female	47.0	1	0	61.1750	S	1	1.000000
15	907	2	del Carlo, Mrs. Sebastiano (Argenia Genovesi)	female	24.0	1	0	27.7208	C	1	1.000000
16	908	2	Keane, Mr. Daniel	male	35.0	0	0	12.3500	Q	0	0.025000
17	909	3	Assaf, Mr. Gerios	male	21.0	0	0	7.2250	C	1	0.800000
18	910	3	Ilmakangas, Miss. Ida Livija	female	27.0	1	0	7.9250	S	1	0.600000
19	911	3	Assaf Khalil, Mrs. Mariana (Miriam")"	female	45.0	0	0	7.2250	C	0	0.400000
20	912	1	Rothschild, Mr. Martin	male	55.0	1	0	59.4000	C	0	0.200000
21	913	3	Olsen, Master. Artur Karl	male	9.0	0	1	3.1708	S	0	0.200000
22	914	1	Flegenheim, Mrs. Alfred (Antoinette)	female	NaN	0	0	31.6833	S	1	0.900000
23	915	1	Williams, Mr. Richard Norris II	male	21.0	0	1	61.3792	C	0	0.500000
24	916	1	Ryerson, Mrs. Arthur Larned (Emily Maria Borie)	female	48.0	1	3	262.3750	C	1	1.000000
25	917	3	Robins, Mr. Alexander A	male	50.0	1	0	14.5000	S	0	0.000000
26	918	1	Ostby, Miss. Helene Ragnhild	female	22.0	0	1	61.9792	C	1	1.000000
27	919	3	Daher, Mr. Shedid	male	22.5	0	0	7.2250	C	1	0.800000
28	920	1	Brady, Mr. John Bertram	male	41.0	0	0	30.5000	S	1	0.800000
29	921	3	Samaan, Mr. Elias	male	NaN	2	0	21.6792	C	0	0.000000
...	...	...	...	...	...	...	...	...	...	...	...
388	1280	3	Canavan, Mr. Patrick	male	21.0	0	0	7.7500	Q	0	0.000000
389	1281	3	Palsson, Master. Paul Folke	male	6.0	3	1	21.0750	S	0	0.400000
390	1282	1	Payne, Mr. Vivian Ponsonby	male	23.0	0	0	93.5000	S	0	0.100000
391	1283	1	Lines, Mrs. Ernest H (Elizabeth Lindsey James)	female	51.0	0	1	39.4000	S	1	0.900000
392	1284	3	Abbott, Master. Eugene Joseph	male	13.0	0	2	20.2500	S	0	0.100000
393	1285	2	Gilbert, Mr. William	male	47.0	0	0	10.5000	S	0	0.000000
394	1286	3	Kink-Heilmann, Mr. Anton	male	29.0	3	1	22.0250	S	0	0.300000
395	1287	1	Smith, Mrs. Lucien Philip (Mary Eloise Hughes)	female	18.0	1	0	60.0000	S	1	1.000000
396	1288	3	Colbert, Mr. Patrick	male	24.0	0	0	7.2500	Q	0	0.000000
397	1289	1	Frolicher-Stehli, Mrs. Maxmillian (Margaretha ...	female	48.0	1	1	79.2000	C	1	1.000000
398	1290	3	Larsson-Rondberg, Mr. Edvard A	male	22.0	0	0	7.7750	S	0	0.000000
399	1291	3	Conlon, Mr. Thomas Henry	male	31.0	0	0	7.7333	Q	0	0.000000
400	1292	1	Bonnell, Miss. Caroline	female	30.0	0	0	164.8667	S	1	1.000000
401	1293	2	Gale, Mr. Harry	male	38.0	1	0	21.0000	S	0	0.000000
402	1294	1	Gibson, Miss. Dorothy Winifred	female	22.0	0	1	59.4000	C	1	1.000000
403	1295	1	Carrau, Mr. Jose Pedro	male	17.0	0	0	47.1000	S	0	0.100000
404	1296	1	Frauenthal, Mr. Isaac Gerald	male	43.0	1	0	27.7208	C	0	0.100000
405	1297	2	Nourney, Mr. Alfred (Baron von Drachstedt")"	male	20.0	0	0	13.8625	C	0	0.300000
406	1298	2	Ware, Mr. William Jeffery	male	23.0	1	0	10.5000	S	0	0.000000
407	1299	1	Widener, Mr. George Dunton	male	50.0	1	1	211.5000	C	0	0.500000
408	1300	3	Riordan, Miss. Johanna Hannah""	female	NaN	0	0	7.7208	Q	1	1.000000
409	1301	3	Peacock, Miss. Treasteall	female	3.0	1	1	13.7750	S	1	0.800000
410	1302	3	Naughton, Miss. Hannah	female	NaN	0	0	7.7500	Q	1	0.929808
411	1303	1	Minahan, Mrs. William Edward (Lillian E Thorpe)	female	37.0	1	0	90.0000	Q	1	1.000000
412	1304	3	Henriksson, Miss. Jenny Lovisa	female	28.0	0	0	7.7750	S	1	0.533333
413	1305	3	Spector, Mr. Woolf	male	NaN	0	0	8.0500	S	0	0.200000
414	1306	1	Oliva y Ocana, Dona. Fermina	female	39.0	0	0	108.9000	C	1	1.000000
415	1307	3	Saether, Mr. Simon Sivertsen	male	38.5	0	0	7.2500	S	0	0.000000
416	1308	3	Ware, Mr. Frederick	male	NaN	0	0	8.0500	S	0	0.200000
417	1309	3	Peter, Master. Michael J	male	NaN	1	1	22.3583	C	0	0.100000

418 rows × 11 columns

戻る