[머신러닝] 논리회귀 실습( 이진논리회귀, 다항논리회귀 )

이진논리회귀 실습

1. 데이터다운받기

import os
os.environ['KAGGLE_USERNAME'] = 'username' # username
os.environ['KAGGLE_KEY'] = 'key' # key

!kaggle datasets download -d heptapod/titanic
!unzip titanic.zip

2. 필요한 패키지 임포트하기

from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense
from tensorflow.keras.optimizers import Adam, SGD
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt 
import seaborn as sns
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import StandardScaler

3. 데이터 로딩하기

df = pd.read_csv('train_and_test2.csv')

4. 전처리하기

df = pd.read_csv('train_and_test2.csv', usecols=[
  'Age', # 나이
  'Fare', # 승차 요금
  'Sex', # 성별
  'sibsp', # 타이타닉에 탑승한 형제자매, 배우자의 수
  'Parch', # 타이타니게 탑승한 부모, 자식의 수
  'Pclass', # 티켓 등급 (1, 2, 3등석)
  'Embarked', # 탑승국
  '2urvived' # 생존 여부 (0: 사망, 1: 생존)
])

df = df.dropna()

x_data = df.drop(columns=['2urvived'], axis=1)
x_data = x_data.astype(np.float32)

y_data = df[['2urvived']]
y_data = y_data.astype(np.float32)

scaler = StandardScaler()
x_data_scaled = scaler.fit_transform(x_data)

x_train, x_val, y_train, y_val = train_test_split(x_data, y_data, test_size=0.2, random_state=2021)

5. 모델학습하기

model = Sequential([
  Dense(1, activation='sigmoid')
])

model.compile(loss='binary_crossentropy', optimizer=Adam(lr=0.01), metrics=['acc'])

model.fit(
    x_train,
    y_train,
    validation_data=(x_val, y_val), # 검증 데이터를 넣어주면 한 epoch이 끝날때마다 자동으로 검증
    epochs=20 # epochs 복수형으로 쓰기!
)

 

다항논리회귀 실습

import os
os.environ['KAGGLE_USERNAME'] = 'username' # username
os.environ['KAGGLE_KEY'] = 'key' # key
!kaggle datasets download -d brynja/wineuci
!unzip wineuci.zip

from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense
from tensorflow.keras.optimizers import Adam, SGD
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt 
import seaborn as sns
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import StandardScaler
from sklearn.preprocessing import OneHotEncoder

df = pd.read_csv('Wine.csv')

전처리

df = pd.read_csv('Wine.csv', names=[
  'name'
  ,'alcohol'
  ,'malicAcid'
  ,'ash'
  ,'ashalcalinity'
  ,'magnesium'
  ,'totalPhenols'
  ,'flavanoids'
  ,'nonFlavanoidPhenols'
  ,'proanthocyanins'
  ,'colorIntensity'
  ,'hue'
  ,'od280_od315'
  ,'proline'
])

x_data = df.drop(columns=['name'], axis=1)
x_data = x_data.astype(np.float32)

y_data = df[['name']]
y_data = y_data.astype(np.float32)

scaler = StandardScaler()
x_data_scaled = scaler.fit_transform(x_data)

encoder = OneHotEncoder()
y_data_encoded = encoder.fit_transform(y_data).toarray()

x_train, x_val, y_train, y_val = train_test_split(x_data_scaled, y_data_encoded, test_size=0.2, random_state=2021)

모델학습

model = Sequential([
  Dense(3, activation='softmax')
])

model.compile(loss='categorical_crossentropy', optimizer=Adam(lr=0.02), metrics=['acc'])

model.fit(
    x_train,
    y_train,
    validation_data=(x_val, y_val), # 검증 데이터를 넣어주면 한 epoch이 끝날때마다 자동으로 검증
    epochs=20 # epochs 복수형으로 쓰기!
)