Melakukan Prediksi Menggunakan Machine Learning

Pada tulisan sebelumnya telah dilakukan analisis dengan Korelasi Pearson untuk mengetahui fitur yang paling berpengaruh pada dataset :

DATASET : https://archive.ics.uci.edu/ml/datasets/Energy+efficiency

https://archive.ics.uci.edu/ml/machine-learning-databases/00242/ENB2012_data.xlsx

NAMA FILE : ENB2012_data.xlsx

768 sampel, Setiap sampel terdiri dari 8 fitur numerik yang terdiri dari faktor-faktor seperti luas permukaan, tinggi plafon, dan jenis kaca yang digunakan.

1. Kolom X1 : compactness relatif dari bangunan (Relative Compactness), yang merupakan rasio antara volume bangunan dan luas permukaan eksterior bangunan.
2. Kolom X2 : permukaan (Surface Area) dari bangunan.
3. Kolom : luas dinding (Wall Area) dari bangunan.
4. Kolom X4 : luas atap (Roof Area) dari bangunan.
5. Kolom X5: tinggi keseluruhan (Overall Height) dari bangunan.
6. Kolom X6 : orientasi (Orientation) bangunan. Kolom ini mengindikasikan arah pandang bangunan terhadap matahari, diukur dalam derajat dari utara dalam arah jarum jam.
7. Kolom X7 : luas kaca (Glazing Area) dari bangunan.
8. Kolom X8 : distribusi luas kaca (Glazing Area Distribution) pada bangunan.
9. Kolom Y1 : koefisien transmisi panas (Heat Transmission Coefficient) dari bangunan, yang mengukur seberapa baik bangunan dapat menahan panas dari luar ke dalam.
10. Kolom Y2 : kehilangan energi pendinginan (Cooling Load) dari bangunan, yang mengukur jumlah energi yang dibutuhkan untuk mendinginkan ruangan dalam bangunan.

Regresi Linier Berganda

Pada data set tersebut akan dilakukan regresi berganda untuk memprediksi nilai Y1 dan Y2, data yang akan diuji akan diambil secara acak sebanyak 20% dari data data latih yang disediakan (dataset)

import pandas as pd

import numpy as np

import matplotlib.pyplot as plt

from sklearn.linear_model import LinearRegression

from sklearn.preprocessing import PolynomialFeatures

from sklearn.metrics import r2_score, mean_squared_error

# Membaca dataset

url = "https://archive.ics.uci.edu/ml/machine-learning-databases/00242/ENB2012_data.xlsx"

df = pd.read_excel(url)

# Mengganti nama kolom pada dataset

df = df.rename(columns={'X1': 'relative compactness',

                        'X2': 'surface area',

                        'X3': 'wall area',

                        'X4': 'roof area',

                        'X5': 'overall height',

                        'X6': 'orientation',

                        'X7': 'glazing area',

                        'X8': 'glazing area distribution',

                        'Y1': 'heating load',

                        'Y2': 'cooling load'})

# Mengambil variabel yang akan digunakan dalam analisis

X = df[['roof area', 'overall height']]

y1 = df['heating load']

y2 = df['cooling load']

# Membagi dataset menjadi data latih dan data uji

from sklearn.model_selection import train_test_split

X_train, X_test, y1_train, y1_test, y2_train, y2_test = train_test_split(X, y1, y2, test_size=0.2, random_state=0)

# Melakukan regresi polinomial pada data latih

poly = PolynomialFeatures(degree=2)

X_poly_train = poly.fit_transform(X_train)

X_poly_test = poly.transform(X_test)

regressor1 = LinearRegression()

regressor1.fit(X_poly_train, y1_train)

regressor2 = LinearRegression()

regressor2.fit(X_poly_train, y2_train)

# Memprediksi nilai y berdasarkan data uji

y1_pred = regressor1.predict(X_poly_test)

y2_pred = regressor2.predict(X_poly_test)

# Menampilkan persamaan regresi

print("Persamaan Regresi Heating Load: Y = {} + {}*X1 + {}*X2 + {}*X1^2 + {}*X2^2 + {}*X1*X2".format(

    regressor1.intercept_, regressor1.coef_[0], regressor1.coef_[1], regressor1.coef_[2], regressor1.coef_[3], regressor1.coef_[4]))

print("Persamaan Regresi Cooling Load: Y = {} + {}*X1 + {}*X2 + {}*X1^2 + {}*X2^2 + {}*X1*X2".format(

    regressor2.intercept_, regressor2.coef_[0], regressor2.coef_[1], regressor2.coef_[2], regressor2.coef_[3], regressor2.coef_[4]))

# Menampilkan plot kurva regresi

fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(10,5))

ax1.scatter(X_test['roof area'], y1_test, color='red')

ax1.plot(X_test['roof area'], y1_pred, color='blue', linewidth=2)

ax1.set_xlabel('roof area')

ax1.set_ylabel('Heating Load')

ax2.scatter(X_test['roof area'], y2_test, color='blue')

ax2.plot(X_test['roof area'], y2_pred, color='red', linewidth=2)

ax2.set_xlabel('roof area')

ax2.set_ylabel('Cooling Load')

# Menampilkan plot kurva regresi

fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(10,5))

ax1.scatter(X_test['overall height'], y1_test, color='red')

ax1.plot(X_test['overall height'], y1_pred, color='blue', linewidth=2)

ax1.set_xlabel('overall height')

ax1.set_ylabel('Heating Load')

ax2.scatter(X_test['overall height'], y2_test, color='blue')

ax2.plot(X_test['overall height'], y2_pred, color='red', linewidth=2)

ax2.set_xlabel('overall height')

ax2.set_ylabel('Cooling Load')

# Menampilkan kolom aktual dan hasil prediksi Y1

result1 = pd.DataFrame({'Actual Y1': y1_test, 'Predicted Y1': y1_pred})

print(result1)

# Menampilkan kolom aktual dan hasil prediksi Y2

result2 = pd.DataFrame({'Actual Y2': y2_test, 'Predicted Y2': y2_pred})

print(result2)

# Menyimpan kolom aktual dan hasil prediksi Y1 dan Y2 ke file excel

with pd.ExcelWriter('predicted_results.xlsx') as writer:

    result1.to_excel(writer, sheet_name='Y1')

    result2.to_excel(writer, sheet_name='Y2')

# Menghitung MAE

from sklearn.metrics import mean_absolute_error

mae1 = mean_absolute_error(y1_test, y1_pred)

mae2 = mean_absolute_error(y2_test, y2_pred)

# Menampilkan nilai R-squared, MSE, dan MAE

print("Heating Load:")

print("R-squared:", r2_score(y1_test, y1_pred))

print("MSE:", mean_squared_error(y1_test, y1_pred))

print("MAE:", mae1)

print("Cooling Load:")

print("R-squared:", r2_score(y2_test, y2_pred))

print("MSE:", mean_squared_error(y2_test, y2_pred))

print("MAE:", mae2)

# Menyimpan nilai R-squared, MSE, dan MAE ke file excel

data = {'R-squared': [r2_score(y1_test, y1_pred), r2_score(y2_test, y2_pred)],

        'MSE': [mean_squared_error(y1_test, y1_pred), mean_squared_error(y2_test, y2_pred)],

        'MAE': [mae1, mae2]}

df = pd.DataFrame(data, index=['Heating Load', 'Cooling Load'])

df.to_excel(writer, sheet_name='Metrics')

HASIL

-----------------------------------------------------------------------------

import pandas as pd

import numpy as np

import matplotlib.pyplot as plt

from sklearn.linear_model import LinearRegression

from sklearn.preprocessing import PolynomialFeatures

from sklearn.metrics import r2_score, mean_squared_error

# Membaca dataset

url = "https://archive.ics.uci.edu/ml/machine-learning-databases/00242/ENB2012_data.xlsx"

df = pd.read_excel(url)

# Mengganti nama kolom pada dataset

df = df.rename(columns={'X1': 'relative compactness',

                        'X2': 'surface area',

                        'X3': 'wall area',

                        'X4': 'roof area',

                        'X5': 'overall height',

                        'X6': 'orientation',

                        'X7': 'glazing area',

                        'X8': 'glazing area distribution',

                        'Y1': 'heating load',

                        'Y2': 'cooling load'})

# Mengambil variabel yang akan digunakan dalam analisis

X = df[['roof area', 'overall height']]

y1 = df['heating load']

y2 = df['cooling load']

# Membagi dataset menjadi data latih dan data uji

from sklearn.model_selection import train_test_split

X_train, X_test, y1_train, y1_test, y2_train, y2_test = train_test_split(X, y1, y2, test_size=0.2, random_state=0)

# Melakukan regresi polinomial pada data latih

poly = PolynomialFeatures(degree=2)

X_poly_train = poly.fit_transform(X_train)

X_poly_test = poly.transform(X_test)

regressor1 = LinearRegression()

regressor1.fit(X_poly_train, y1_train)

regressor2 = LinearRegression()

regressor2.fit(X_poly_train, y2_train)

# Memprediksi nilai y berdasarkan data uji

y1_pred = regressor1.predict(X_poly_test)

y2_pred = regressor2.predict(X_poly_test)

# Menampilkan persamaan regresi

print("Persamaan Regresi Heating Load: Y = {} + {}*X1 + {}*X2 + {}*X1^2 + {}*X2^2 + {}*X1*X2".format(

    regressor1.intercept_, regressor1.coef_[0], regressor1.coef_[1], regressor1.coef_[2], regressor1.coef_[3], regressor1.coef_[4]))

print("Persamaan Regresi Cooling Load: Y = {} + {}*X1 + {}*X2 + {}*X1^2 + {}*X2^2 + {}*X1*X2".format(

    regressor2.intercept_, regressor2.coef_[0], regressor2.coef_[1], regressor2.coef_[2], regressor2.coef_[3], regressor2.coef_[4]))

# Menampilkan plot kurva regresi

fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(10,5))

ax1.scatter(X_test['roof area'], y1_test, color='red')

ax1.plot(X_test['roof area'], y1_pred, color='blue', linewidth=2)

ax1.set_xlabel('roof area')

ax1.set_ylabel('Heating Load')

ax2.scatter(X_test['roof area'], y2_test, color='blue')

ax2.plot(X_test['roof area'], y2_pred, color='red', linewidth=2)

ax2.set_xlabel('roof area')

ax2.set_ylabel('Cooling Load')

# Menampilkan plot kurva regresi

fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(10,5))

ax1.scatter(X_test['overall height'], y1_test, color='red')

ax1.plot(X_test['overall height'], y1_pred, color='blue', linewidth=2)

ax1.set_xlabel('overall height')

ax1.set_ylabel('Heating Load')

ax2.scatter(X_test['overall height'], y2_test, color='blue')

ax2.plot(X_test['overall height'], y2_pred, color='red', linewidth=2)

ax2.set_xlabel('overall height')

ax2.set_ylabel('Cooling Load')

# Menampilkan kolom aktual dan hasil prediksi Y1

result1 = pd.DataFrame({'Actual Y1': y1_test, 'Predicted Y1': y1_pred})

print(result1)

# Menampilkan kolom aktual dan hasil prediksi Y2

result2 = pd.DataFrame({'Actual Y2': y2_test, 'Predicted Y2': y2_pred})

print(result2)

# Menyimpan kolom aktual dan hasil prediksi Y1 dan Y2 ke file excel

with pd.ExcelWriter('predicted_results.xlsx') as writer:

    result1.to_excel(writer, sheet_name='Y1')

    result2.to_excel(writer, sheet_name='Y2')

# Menghitung MAE

from sklearn.metrics import mean_absolute_error

mae1 = mean_absolute_error(y1_test, y1_pred)

mae2 = mean_absolute_error(y2_test, y2_pred)

# Menampilkan nilai R-squared, MSE, dan MAE

print("Heating Load:")

print("R-squared:", r2_score(y1_test, y1_pred))

print("MSE:", mean_squared_error(y1_test, y1_pred))

print("MAE:", mae1)

print("Cooling Load:")

print("R-squared:", r2_score(y2_test, y2_pred))

print("MSE:", mean_squared_error(y2_test, y2_pred))

print("MAE:", mae2)

# Menyimpan nilai R-squared, MSE, dan MAE ke file excel

data = {'R-squared': [r2_score(y1_test, y1_pred), r2_score(y2_test, y2_pred)],

        'MSE': [mean_squared_error(y1_test, y1_pred), mean_squared_error(y2_test, y2_pred)],

        'MAE': [mae1, mae2]}

df = pd.DataFrame(data, index=['Heating Load', 'Cooling Load'])

df.to_excel(writer, sheet_name='Metrics')