import pandas as pd
import numpy as np
import [Link] as plt
from [Link] import LabelEncoder, MinMaxScaler
from sklearn.model_selection import train_test_split
from [Link] import SVR
from [Link] import RandomForestRegressor
from [Link] import mean_absolute_error, mean_squared_error
import ipywidgets as widgets
from [Link] import display, clear_output
import warnings
# Suppress warnings for cleaner output
[Link]("ignore")
# Load the dataset
ipl = pd.read_csv('/content/ipl_data.csv')
# Drop unimportant features
df = [Link](['date', 'runs', 'wickets', 'overs', 'runs_last_5', 'wickets_last_5', 'mid
# Split into features (X) and target (y)
X = [Link](['total'], axis=1)
y = df['total']
# Initialize LabelEncoders for categorical features
venue_encoder = LabelEncoder()
batting_team_encoder = LabelEncoder()
bowling_team_encoder = LabelEncoder()
batsman_encoder = LabelEncoder()
bowler_encoder = LabelEncoder()
# Encode categorical features
X['venue'] = venue_encoder.fit_transform(X['venue'])
X['bat_team'] = batting_team_encoder.fit_transform(X['bat_team'])
X['bowl_team'] = bowling_team_encoder.fit_transform(X['bowl_team'])
X['batsman'] = batsman_encoder.fit_transform(X['batsman'])
X['bowler'] = bowler_encoder.fit_transform(X['bowler'])
# Split data into training and testing sets
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=42)
# Scale the features
scaler = MinMaxScaler()
X_train_scaled = scaler.fit_transform(X_train)
X_test_scaled = [Link](X_test)
# Initialize models
svr_model = SVR(kernel='rbf', C=100, epsilon=0.1) # Support Vector Regressor with RBF ke
rf_model = RandomForestRegressor(n_estimators=100, random_state=42) # Random Forest Regr
# Train both models
svr_model.fit(X_train_scaled, y_train)
rf_model.fit(X_train_scaled, y_train)
▾ RandomForestRegressor i ?
RandomForestRegressor(random_state=42)
# Make predictions
svr_predictions = svr_model.predict(X_test_scaled)
rf_predictions = rf_model.predict(X_test_scaled)
# Evaluate models
svr_mae = mean_absolute_error(y_test, svr_predictions)
svr_mse = mean_squared_error(y_test, svr_predictions)
rf_mae = mean_absolute_error(y_test, rf_predictions)
rf_mse = mean_squared_error(y_test, rf_predictions)
# Print evaluation metrics
print("Support Vector Regressor:")
print(f"Mean Absolute Error: {svr_mae:.2f}")
print(f"Mean Squared Error: {svr_mse:.2f}")
print("\nRandom Forest Regressor:")
print(f"Mean Absolute Error: {rf_mae:.2f}")
print(f"Mean Squared Error: {rf_mse:.2f}")
Support Vector Regressor:
Mean Absolute Error: 19.21
Mean Squared Error: 715.16
Random Forest Regressor:
Mean Absolute Error: 2.14
Mean Squared Error: 54.46
# Plot actual vs predicted scores for both models
[Link](figsize=(12, 5))
# SVR Plot
[Link](1, 2, 1)
[Link](y_test, svr_predictions, alpha=0.5)
[Link]([y_test.min(), y_test.max()], [y_test.min(), y_test.max()], 'r--', lw=2)
[Link]('Actual Scores')
[Link]('Predicted Scores')
[Link]('SVR: Actual vs Predicted Scores')
# Random Forest Plot
[Link](1, 2, 2)
[Link](y_test, rf_predictions, alpha=0.5)
[Link]([y_test.min(), y_test.max()], [y_test.min(), y_test.max()], 'r--', lw=2)
[Link]('Actual Scores')
[Link]('Predicted Scores')
[Link]('Random Forest: Actual vs Predicted Scores')
plt.tight_layout()
[Link]()
# Create interactive widgets for prediction
venue = [Link](options=df['venue'].unique().tolist(), description='Select Venue
batting_team = [Link](options=df['bat_team'].unique().tolist(), description='Se
bowling_team = [Link](options=df['bowl_team'].unique().tolist(), description='S
batsman = [Link](options=df['batsman'].unique().tolist(), description='Select B
bowler = [Link](options=df['bowler'].unique().tolist(), description='Select Bow
model_choice = [Link](options=['SVR', 'Random Forest'], description='Select Mod
predict_button = [Link](description="Predict Score")
output = [Link]()
# Define prediction function
def predict_score(b):
with output:
clear_output() # Clear previous output
try:
# Encode user inputs
encoded_venue = venue_encoder.transform([[Link]])[0]
encoded_batting_team = batting_team_encoder.transform([batting_team.value])[0
encoded_bowling_team = bowling_team_encoder.transform([bowling_team.value])[0
encoded_batsman = batsman_encoder.transform([[Link]])[0]
encoded_bowler = bowler_encoder.transform([[Link]])[0]
# Create input array
input_data = [Link]([[encoded_venue, encoded_batting_team, encoded_bowling_
input_scaled = [Link](input_data) # Scale input
# Select model for prediction
if model_choice.value == 'SVR':
predicted_score = svr_model.predict(input_scaled)[0]
model_name = 'SVR'
else:
predicted_score = rf_model.predict(input_scaled)[0]
model_name = 'Random Forest'
print(f"Predicted Score ({model_name}): {int(predicted_score)}")
except Exception as e:
print(f"Error in prediction: {str(e)}")
# Link button to prediction function
predict_button.on_click(predict_score)
# Display widgets
display(venue, batting_team, bowling_team, batsman, bowler, model_choice, predict_button
Select Venue: M Chinnaswamy Stadium
Select Battin… Chennai Super Kings
Select Bowlin… Royal Challengers Bangalore
Select Batsm… MS Dhoni
Select Bowler: PP Chawla
Select Model: Random Forest
Predict Score
Predicted Score (Random Forest): 175
