import numpy as np from sklearn.svm import SVC from sklearn.datasets import make_classification from sklearn.model_selection import train_test_split from sklearn.preprocessing import StandardScaler # 線形分離可能なデータを生成 X, y = make_classification( n_samples=200, n_features=2, n_informative=2, n_redundant=0, n_clusters_per_class=1, class_sep=1.5, random_state=42 ) # データ分割 X_train, X_test, y_train, y_test = train_test_split( X, y, test_size=0.3, random_state=42 ) # スケーリング（重要！） scaler = StandardScaler() X_train_scaled = scaler.fit_transform(X_train) X_test_scaled = scaler.transform(X_test) # 線形SVM svm_linear = SVC(kernel=’linear’, C=1.0, random_state=42) svm_linear.fit(X_train_scaled, y_train) print(“=== 線形SVMの結果 ===”) print(f”訓練精度: {svm_linear.score(X_train_scaled, y_train):.4f}”) print(f”テスト精度: {svm_linear.score(X_test_scaled, y_test):.4f}”) print(f”\nサポートベクターの数: {len(svm_linear.support_vectors_)}”) print(f”全訓練データに対する割合: {len(svm_linear.support_vectors_) / len(X_train) * 100:.1f}%”)

from sklearn.datasets import make_circles, make_moons # 同心円データを生成 X_circles, y_circles = make_circles(n_samples=200, noise=0.1, factor=0.5, random_state=42) # スケーリング scaler = StandardScaler() X_scaled = scaler.fit_transform(X_circles) # 異なるカーネルを試す kernels = [‘linear’, ‘rbf’, ‘poly’] print(“=== カーネル別の精度（同心円データ） ===”) for kernel in kernels: svm = SVC(kernel=kernel, gamma=’auto’, random_state=42) svm.fit(X_scaled, y_circles) accuracy = svm.score(X_scaled, y_circles) print(f”{kernel:10s}: {accuracy:.4f}”)

# サンプルデータ X_sample, y_sample = make_classification( n_samples=100, n_features=2, n_redundant=0, n_informative=2, random_state=42 ) scaler = StandardScaler() X_sample_scaled = scaler.fit_transform(X_sample) # データ分割 X_train, X_test, y_train, y_test = train_test_split( X_sample_scaled, y_sample, test_size=0.3, random_state=42 ) # 異なるCとgammaを試す C_values = [0.1, 1, 10, 100] gamma_values = [0.01, 0.1, 1, 10] print(“=== Cとgammaの組み合わせによる精度 ===”) print(f”{‘C’:<10} {'gamma':<10} {'Train':<10} {'Test':<10}") print("-" * 40) for C in C_values: for gamma in gamma_values: svm = SVC(kernel='rbf', C=C, gamma=gamma, random_state=42) svm.fit(X_train, y_train) train_acc = svm.score(X_train, y_train) test_acc = svm.score(X_test, y_test) print(f"{C:<10} {gamma:<10} {train_acc:<10.4f} {test_acc:<10.4f}")

from sklearn.model_selection import GridSearchCV # パラメータグリッド param_grid = { ‘C’: [0.1, 1, 10, 100], ‘gamma’: [0.001, 0.01, 0.1, 1], ‘kernel’: [‘rbf’] } # GridSearch grid_search = GridSearchCV( SVC(random_state=42), param_grid, cv=5, scoring=’accuracy’, n_jobs=-1 ) grid_search.fit(X_train, y_train) print(“=== GridSearchCVの結果 ===”) print(f”最良パラメータ: {grid_search.best_params_}”) print(f”最良CVスコア: {grid_search.best_score_:.4f}”) print(f”テストスコア: {grid_search.score(X_test, y_test):.4f}”)

from sklearn.datasets import load_digits from sklearn.linear_model import LogisticRegression from sklearn.tree import DecisionTreeClassifier from sklearn.ensemble import RandomForestClassifier import pandas as pd # 手書き数字データセット digits = load_digits() X, y = digits.data, digits.target print(f”サンプル数: {len(X)}, 特徴量数: {X.shape[1]}, クラス数: {len(set(y))}”) # データ分割 X_train, X_test, y_train, y_test = train_test_split( X, y, test_size=0.3, random_state=42, stratify=y ) # スケーリング scaler = StandardScaler() X_train_scaled = scaler.fit_transform(X_train) X_test_scaled = scaler.transform(X_test) # 複数のモデルを比較 models = { ‘Logistic Regression’: LogisticRegression(max_iter=1000, random_state=42), ‘Decision Tree’: DecisionTreeClassifier(max_depth=10, random_state=42), ‘Random Forest’: RandomForestClassifier(n_estimators=100, random_state=42), ‘SVM (linear)’: SVC(kernel=’linear’, random_state=42), ‘SVM (RBF)’: SVC(kernel=’rbf’, gamma=’scale’, random_state=42) } print(“\n=== モデルの比較 ===”) for name, model in models.items(): model.fit(X_train_scaled, y_train) train_acc = model.score(X_train_scaled, y_train) test_acc = model.score(X_test_scaled, y_test) print(f”{name:25s}: Train={train_acc:.4f}, Test={test_acc:.4f}”)