AutoML

import autosklearn.classification
import pandas as pd
from sklearn.model_selection import train_test_split
from sklearn.metrics import accuracy_score

# Load and prepare data
df = pd.read_csv('dataset.csv')
X = df.drop('target', axis=1)
y = df['target']
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

# AutoML with time budget
automl = autosklearn.classification.AutoSklearnClassifier(
    time_left_for_this_task=3600,  # 1 hour
    per_run_time_limit=300,        # 5 minutes per model
    memory_limit=8192,             # 8GB RAM limit
    ensemble_size=50,              # Ensemble size
    initial_configurations_via_metalearning=25,
    seed=42
)

# Fit AutoML
automl.fit(X_train, y_train)

# Get predictions and statistics
y_pred = automl.predict(X_test)
accuracy = accuracy_score(y_test, y_pred)
print(f"AutoML Accuracy: {accuracy:.4f}")

# Show ensemble statistics
print(f"Ensemble size: {len(automl.get_models_with_weights())}")
print(automl.sprint_statistics())

from tpot import TPOTClassifier
from sklearn.datasets import load_digits
from sklearn.model_selection import train_test_split

# Load dataset
digits = load_digits()
X_train, X_test, y_train, y_test = train_test_split(
    digits.data, digits.target, test_size=0.25, random_state=42
)

# Configure TPOT
tpot = TPOTClassifier(
    generations=50,         # Number of iterations
    population_size=50,     # Number of models per generation
    cv=5,                  # Cross-validation folds
    random_state=42,
    verbosity=2,
    max_time_mins=60,      # Maximum optimization time
    early_stop=10          # Early stopping
)

# Fit and evolve pipelines
tpot.fit(X_train, y_train)

# Evaluate best pipeline
accuracy = tpot.score(X_test, y_test)
print(f"Best pipeline accuracy: {accuracy:.4f}")

# Export optimized pipeline
tpot.export('tpot_optimized_pipeline.py')

# Show best pipeline
print("Best pipeline:")
print(tpot.fitted_pipeline_)

import autokeras as ak
import numpy as np
import tensorflow as tf
from tensorflow.keras.datasets import cifar10

# Load CIFAR-10 dataset
(x_train, y_train), (x_test, y_test) = cifar10.load_data()

# Initialize AutoKeras classifier
clf = ak.ImageClassifier(
    max_trials=20,        # Maximum number of models to try
    epochs=10,            # Training epochs per trial
    directory='automl_logs',
    overwrite=True,
    seed=42
)

# Search for best architecture
clf.fit(x_train, y_train, 
        validation_split=0.15,
        verbose=1)

# Evaluate on test set
accuracy = clf.evaluate(x_test, y_test, verbose=0)[1]
print(f"Best model accuracy: {accuracy:.4f}")

# Get the best model
best_model = clf.export_model()
print("Best architecture summary:")
best_model.summary()

# Save the best model
best_model.save('best_automl_model.h5')