import pandas as pd
import numpy as np
from sklearn.model_selection import train_test_split, GridSearchCV
from sklearn.metrics import mean_squared_error, mean_absolute_error,
r2_score
from sklearn.tree import DecisionTreeRegressor
from sklearn.ensemble import RandomForestRegressor
from xgboost import XGBRegressor
from sklearn.preprocessing import StandardScaler
# For ANN
import tensorflow as tf
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense
from tensorflow.keras.wrappers.scikit_learn import KerasRegressor
# Load dataset
df = pd.read_csv("Imbalanced_PV_Fault_Dataset.csv")
X = df.drop(columns=["Feature_0", "Fault_Flag"])
y = df["Feature_0"]
# Train-test split
X_train, X_test, y_train, y_test = train_test_split(X, y,
test_size=0.2, random_state=42)
# Scaling
scaler = StandardScaler()
X_train_scaled = scaler.fit_transform(X_train)
X_test_scaled = scaler.transform(X_test)
# ----------- Recursive Least Squares -----------
class RecursiveLeastSquares:
def __init__(self, num_features, lambda_factor=0.99, delta=1.0):
self.num_features = num_features
self.lambda_factor = lambda_factor
self.P = np.eye(num_features) * delta
self.theta = np.zeros(num_features)
def update(self, X, y):
X = np.array(X).reshape(-1, 1)
y = np.array(y)
K = self.P @ X / (self.lambda_factor + X.T @ self.P @ X)
self.theta += (y - X.T @ self.theta) * K.flatten()
self.P = (self.P - K @ X.T @ self.P) / self.lambda_factor
rls = RecursiveLeastSquares(num_features=X.shape[1])
for i in range(len(X_train)):
rls.update(X_train.iloc[i].values, y_train.iloc[i])
y_rls_pred = [np.dot(rls.theta, x) for x in X_test.values]
# ----------- GridSearchCV Models -----------
# Decision Tree
dtr = DecisionTreeRegressor(random_state=42)
dtr_param = {'max_depth': [3, 5, 10, None]}
dtr_grid = GridSearchCV(dtr, dtr_param, cv=5)
dtr_grid.fit(X_train, y_train)
dtr_pred = dtr_grid.best_estimator_.predict(X_test)
# Random Forest
rfr = RandomForestRegressor(random_state=42)
rfr_param = {'n_estimators': [50, 100], 'max_depth': [5, 10, None]}
rfr_grid = GridSearchCV(rfr, rfr_param, cv=5)
rfr_grid.fit(X_train, y_train)
rfr_pred = rfr_grid.best_estimator_.predict(X_test)
# XGBoost
xgb = XGBRegressor(random_state=42, verbosity=0)
xgb_param = {'n_estimators': [50, 100], 'max_depth': [3, 5, 10]}
xgb_grid = GridSearchCV(xgb, xgb_param, cv=5)
xgb_grid.fit(X_train, y_train)
xgb_pred = xgb_grid.best_estimator_.predict(X_test)
# ----------- ANN Model -----------
def build_ann():
model = Sequential()
model.add(Dense(64, input_dim=X_train.shape[1],
activation='relu'))
model.add(Dense(32, activation='relu'))
model.add(Dense(1)) # Output layer
model.compile(optimizer='adam', loss='mse')
return model
ann = KerasRegressor(build_fn=build_ann, epochs=100, batch_size=16,
verbose=0)
ann.fit(X_train_scaled, y_train)
ann_pred = ann.predict(X_test_scaled)
/var/folders/tt/9tcd3n611x1_n7ww1jb91r500000gn/T/
ipykernel_27794/2130103562.py:1: DeprecationWarning: KerasRegressor is
deprecated, use Sci-Keras (https://github.com/adriangb/scikeras)
instead. See https://www.adriangb.com/scikeras/stable/migration.html
for help migrating.
ann = KerasRegressor(build_fn=build_ann, epochs=100, batch_size=16,
verbose=0)
# ----------- Metrics -----------
metrics = {
"Model": ["RLS", "Decision Tree", "Random Forest", "XGBoost",
"ANN"],
"MSE": [
mean_squared_error(y_test, y_rls_pred),
mean_squared_error(y_test, dtr_pred),
mean_squared_error(y_test, rfr_pred),
mean_squared_error(y_test, xgb_pred),
mean_squared_error(y_test, ann_pred),
],
"MAE": [
mean_absolute_error(y_test, y_rls_pred),
mean_absolute_error(y_test, dtr_pred),
mean_absolute_error(y_test, rfr_pred),
mean_absolute_error(y_test, xgb_pred),
mean_absolute_error(y_test, ann_pred),
],
"R² Score": [
r2_score(y_test, y_rls_pred),
r2_score(y_test, dtr_pred),
r2_score(y_test, rfr_pred),
r2_score(y_test, xgb_pred),
r2_score(y_test, ann_pred),
]
}
comparison_df = pd.DataFrame(metrics)
print(comparison_df)
# Optional: print best parameters
print("\nBest Parameters:")
print("DTR:", dtr_grid.best_params_)
print("RFR:", rfr_grid.best_params_)
print("XGB:", xgb_grid.best_params_)
Model MSE MAE R² Score
0 RLS 4.525742e-26 1.691223e-13 1.000000
1 Decision Tree 1.820721e-01 3.104887e-01 0.921079
2 Random Forest 6.800423e-02 1.663037e-01 0.970523
3 XGBoost 4.184689e-02 1.506359e-01 0.981861
4 ANN 2.230879e-04 1.038437e-02 0.999903
Best Parameters:
DTR: {'max_depth': None}
RFR: {'max_depth': None, 'n_estimators': 100}
XGB: {'max_depth': 3, 'n_estimators': 100}
import pandas as pd
# Assuming `comparison_df` already exists from previous code
# and contains columns: "Model", "MSE", "MAE", "R² Score"
# Define output file name
output_file = "comparison_report.xlsx"
# Save to Excel
comparison_df.to_excel(output_file, index=False)
print(f"Comparison report saved as '{output_file}' in your current
working directory.")
Comparison report saved as 'comparison_report.xlsx' in your current
working directory.
