다양한 모델 적용

1. AirQualityUCI 데이터셋

import numpy as np
import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt

air_df = pd.read_csv('/content/drive/MyDrive/KDT/머신러닝과 딥러닝/data/AirQualityUCI.csv')
air_df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 9471 entries, 0 to 9470
Data columns (total 17 columns):
 #   Column         Non-Null Count  Dtype  
---  ------         --------------  -----  
 0   Date           9357 non-null   object 
 1   Time           9357 non-null   object 
 2   CO(GT)         9357 non-null   float64
 3   PT08.S1(CO)    9357 non-null   float64
 4   NMHC(GT)       9357 non-null   float64
 5   C6H6(GT)       9357 non-null   float64
 6   PT08.S2(NMHC)  9357 non-null   float64
 7   NOx(GT)        9357 non-null   float64
 8   PT08.S3(NOx)   9357 non-null   float64
 9   NO2(GT)        9357 non-null   float64
 10  PT08.S4(NO2)   9357 non-null   float64
 11  PT08.S5(O3)    9357 non-null   float64
 12  T              9357 non-null   float64
 13  RH             9357 non-null   float64
 14  AH             9357 non-null   float64
 15  Unnamed: 15    0 non-null      float64
 16  Unnamed: 16    0 non-null      float64
dtypes: float64(15), object(2)
memory usage: 1.2+ MB

• Date: 측정 날짜
• Time: 측정 시간
• CO(GT): 일산화탄소 농도 (mg/m^3)
• PT08.S1(CO): 일산화탄소에 대한 센서 응답
• NMHC(GT): 비메탄 탄화수소 농도 (microg/m^3)
• C6H6(GT): 벤젠 농도 (microg/m^3)
• PT08.S2(NMHC): 탄화수소에 대한 센서 응답
• NOx(GT): 산화 질소 농도 (ppb)
• PT08.S3(NOx): 산화 질소에 대한 센서 응답
• NO2(GT): 이산화질소 농도 (microg/m^3)
• PT08.S4(NO2): 이산화질소에 대한 센서 응답
• PT08.S5(O3): 오존에 대한 센서 응답
• T: 온도 (°C)
• RH: 상대 습도 (%)
• AH: 절대 습도 (g/m^3)

# 필요 없는 열 삭제
air_df.drop(['Unnamed: 15', 'Unnamed: 16'], axis = 1, inplace=True)

# 데이터에 na가 포함되어 있다면 제거
air_df= air_df.dropna()

# Date 컬럼의 데이터를 datetime형식으로 변환
# 날짜-월-년도 형식으로 유지
air_df['Date'] = pd.to_datetime(air_df.Date,format='%d-%m-%Y')
air_df.head()

# Month 파생변수 만들기
# Date 컬럼에서 월을 추출
air_df['month'] = air_df['Date'].dt.month
air_df.head()

# hour 파생변수 만들기
# time에서 시간만 추출
air_df['Time'] = pd.to_datetime(air_df['Time'], format='%H:%M:%S')
air_df['Hour'] = air_df['Time'].dt.hour.astype(int)
# air_df['Hour] = air_df['Time'].str.split(':').str[0].fillna(0).astype(int)
air_df.head()

# Date,Time 컬럼을 제거
air_df.drop(['Date','Time'], axis=1, inplace=True)
air_df.head()

# heatmap을 통해 상관관계를 확인
plt.figure(figsize=(12,12))
sns.heatmap(air_df.corr(), cmap = 'coolwarm', vmax=1, vmin= -1, annot= True)
plt.show()

- StandardScaler로 정규화

from sklearn.preprocessing import StandardScaler
from sklearn.model_selection import train_test_split
from sklearn.metrics import mean_squared_error

ss = StandardScaler()
X = air_df.drop('RH', axis=1)
y = air_df['RH']

Xss = ss.fit_transform(X)

X_train, X_test, y_train, y_test = train_test_split(Xss, y, test_size = 0.2, random_state=2023)

X_train.shape, y_train.shape // ((7485, 14), (7485,))
X_test.shape, y_test.shape // ((1872, 14), (1872,))

2. 모델별 성능 확인을 위한 함수

my_predictions = {}
colors = ['r', 'c', 'm', 'y', 'k', 'khaki', 'teal', 'orchid', 'sandybrown',
          'greenyellow', 'dodgerblue', 'deepskyblue', 'rosybrown', 'firebrick',
          'deeppink', 'crimson', 'salmon', 'darkred', 'olivedrab', 'olive',
          'forestgreen', 'royalblue', 'indigo', 'navy', 'mediumpurple', 'chocolate',
          'gold', 'darkorange', 'seagreen', 'turquoise', 'steelblue', 'slategray',
          'peru', 'midnightblue', 'slateblue', 'dimgray', 'cadetblue', 'tomato']
def plot_predictions(name_, pred, actual):
  df = pd.DataFrame({'prediction': pred, 'actual': y_test})
  df = df.sort_values(by='actual').reset_index(drop=True)
  plt.figure(figsize=(12, 9))
  plt.scatter(df.index, df['prediction'], marker='x', color='r')
  plt.scatter(df.index, df['actual'], alpha=0.7, marker='o', color='black')
  plt.title(name_, fontsize=15)
  plt.legend(['prediction', 'actual'], fontsize=12)
  plt.show()

  # 예측, 실제값
def mse_eval(name_, pred, actual):
  global my_predictions
  global colors
  plot_predictions(name_, pred, actual)
  mse = mean_squared_error(pred, actual)
  my_predictions[name_] = mse
  y_value = sorted(my_predictions.items(), key=lambda x: x[1], reverse=True)
  df = pd.DataFrame(y_value, columns=['model', 'mse'])
  print(df)
  min_ = df['mse'].min() - 10
  max_ = df['mse'].max() + 10
  length = len(df)
  plt.figure(figsize=(10, length))
  ax = plt.subplot()
  ax.set_yticks(np.arange(len(df)))
  ax.set_yticklabels(df['model'], fontsize=15)
  bars = ax.barh(np.arange(len(df)), df['mse'])
  for i, v in enumerate(df['mse']):
    idx = np.random.choice(len(colors))
    bars[i].set_color(colors[idx])
    ax.text(v + 2, i, str(round(v, 3)), color='k', fontsize=15, fontweight='bold')
  plt.title('MSE Error', fontsize=18)
  plt.xlim(min_, max_)
  plt.show()

[ 성능 비교 ]

# Linear Regression
# Decision Tree Regression
# Random Forest Regression
# Suppert Bector Machine
# lightGBM

from sklearn.linear_model import LinearRegression
from sklearn.tree import DecisionTreeRegressor
from sklearn.ensemble import RandomForestRegressor
from sklearn.svm import SVR
from lightgbm import LGBMRegressor

lr = LinearRegression()
dtr = DecisionTreeRegressor(random_state=2023)
rf = RandomForestRegressor()
svm = SVR()
base_model = LGBMRegressor(random_state=2023)

lr.fit(X_train, y_train)
dtr.fit(X_train, y_train)
rf.fit(X_train, y_train)
svm.fit(X_train,y_train)
base_model.fit(X_train,y_train)

pred1 = lr.predict(X_test)
pred2 = dtr.predict(X_test)
pred3 = rf.predict(X_test)
pred4 = svm.predict(X_test)
pred5 = base_model.predict(X_test)

rs1 = np.sqrt(mean_squared_error(y_test, pred1))
rs2 = np.sqrt(mean_squared_error(y_test, pred2))
rs3 = np.sqrt(mean_squared_error(y_test, pred3))
rs4 = np.sqrt(mean_squared_error(y_test, pred4))
rs5 = np.sqrt(mean_squared_error(y_test, pred5))
print(rs1,rs2,rs3,rs4,rs5)
// 7.3316009912232705 1.1906531461890633 0.5864489734872707 19.462400854771953 0.7736864001011999

mse_eval('LinearRegression', pred1, y_test)
mse_eval('DecisionTreeRegressor', pred2, y_test)
mse_eval('RandomForestRegressor', pred3, y_test)
mse_eval('SVR', pred4, y_test)
mse_eval('LGBMRegressor', pred5, y_test)

 

 

 

dict = {'LinearRegression' : rs1, 'DecisionTreeRegressor': rs2, 'RandomForestRegressor':rs3, 'SVR':rs4,'LGBMRegressor':rs5 }

res = [key for key in dict if all(dict[temp] >= dict[key] for temp in dict)]
print(res) // ['RandomForestRegressor']
min = {k: dict[k] for k in dict.keys() & set(res)}
print(min) // {'RandomForestRegressor': 0.5864489734872707}

---

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=2023)
models = {
    "Linear Regression": LinearRegression(),
    "Decision Tree": DecisionTreeRegressor(),
    "Random Forest": RandomForestRegressor(),
    "Gradient Boosting": GradientBoostingRegressor()
}
# Train and evaluate the models
results = {}
for name, model in models.items():
    model.fit(X_train, y_train)
    predictions = model.predict(X_test)
    mse = mean_squared_error(y_test, predictions)
    results[name] = mse
results