ลอจิสติกรีเกรสชัน

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75

from __future__ import annotations

import japanize_matplotlib
import matplotlib.pyplot as plt
import numpy as np
from matplotlib.colors import ListedColormap
from sklearn.datasets import make_classification
from sklearn.linear_model import LogisticRegression
from sklearn.metrics import accuracy_score

def run_logistic_regression_demo(
    n_samples: int = 300,
    random_state: int = 2,
    label_class0: str = "คลาส 0",
    label_class1: str = "คลาส 1",
    label_boundary: str = "เส้นแบ่ง",
    title: str = "เส้นแบ่งจากลอจิสติกรีเกรสชัน",
) -> dict[str, float]:
    """Train logistic regression on a synthetic 2D dataset and visualise the boundary."""
    japanize_matplotlib.japanize()
    X, y = make_classification(
        n_samples=n_samples,
        n_features=2,
        n_redundant=0,
        n_informative=2,
        random_state=random_state,
        n_clusters_per_class=1,
    )

    clf = LogisticRegression()
    clf.fit(X, y)

    accuracy = float(accuracy_score(y, clf.predict(X)))
    coef = clf.coef_[0]
    intercept = float(clf.intercept_[0])

    x1, x2 = X[:, 0], X[:, 1]
    grid_x1, grid_x2 = np.meshgrid(
        np.linspace(x1.min() - 1.0, x1.max() + 1.0, 200),
        np.linspace(x2.min() - 1.0, x2.max() + 1.0, 200),
    )
    grid = np.c_[grid_x1.ravel(), grid_x2.ravel()]
    probs = clf.predict_proba(grid)[:, 1].reshape(grid_x1.shape)

    cmap = ListedColormap(["#aec7e8", "#ffbb78"])
    fig, ax = plt.subplots(figsize=(7, 6))
    contour = ax.contourf(grid_x1, grid_x2, probs, levels=20, cmap=cmap, alpha=0.4)
    ax.contour(grid_x1, grid_x2, probs, levels=[0.5], colors="k", linewidths=1.5)
    ax.scatter(x1[y == 0], x2[y == 0], marker="o", edgecolor="k", label=label_class0)
    ax.scatter(x1[y == 1], x2[y == 1], marker="x", color="k", label=label_class1)
    ax.set_xlabel("คุณลักษณะ 1")
    ax.set_ylabel("คุณลักษณะ 2")
    ax.set_title(title)
    ax.legend(loc="best")
    fig.colorbar(contour, ax=ax, label="P(class = 1)")
    fig.tight_layout()
    plt.show()

    return {
        "accuracy": accuracy,
        "coef_0": float(coef[0]),
        "coef_1": float(coef[1]),
        "intercept": intercept,
    }

metrics = run_logistic_regression_demo(
    label_class0="คลาส 0",
    label_class1="คลาส 1",
    label_boundary="เส้นแบ่ง",
    title="เส้นแบ่งจากลอจิสติกรีเกรสชัน",
)
print(f"ความแม่นยำขณะฝึก: {metrics['accuracy']:.3f}")
print(f"ค่าน้ำหนักคุณลักษณะ 1: {metrics['coef_0']:.3f}")
print(f"ค่าน้ำหนักคุณลักษณะ 2: {metrics['coef_1']:.3f}")
print(f"ไบแอส: {metrics['intercept']:.3f}")