K-yaqin qo'shni (KNN)

import numpy as np
from sklearn.datasets import load_iris
from sklearn.model_selection import train_test_split
from sklearn.neighbors import KNeighborsClassifier
from sklearn.preprocessing import StandardScaler
from sklearn.pipeline import make_pipeline

iris = load_iris()
X, y = iris.data, iris.target

X_o, X_s, y_o, y_s = train_test_split(
    X, y, test_size=0.3, random_state=42, stratify=y
)

model = make_pipeline(StandardScaler(), KNeighborsClassifier(n_neighbors=5))
model.fit(X_o, y_o)

print(f"Aniqlik: {model.score(X_s, y_s):.3f}")

natijalar = []

for k in range(1, 31):
    model = make_pipeline(StandardScaler(), KNeighborsClassifier(n_neighbors=k))
    model.fit(X_o, y_o)

    natijalar.append({
        "k": k,
        "oqitish": model.score(X_o, y_o),
        "sinov": model.score(X_s, y_s),
    })

import pandas as pd
jadval = pd.DataFrame(natijalar)
print(jadval.iloc[[0, 2, 4, 9, 19, 29]].round(3))

from sklearn.model_selection import GridSearchCV

qidiruv = GridSearchCV(
    make_pipeline(StandardScaler(), KNeighborsClassifier()),
    {"kneighborsclassifier__n_neighbors": range(1, 31)},
    cv=5,
    scoring="accuracy",
)
qidiruv.fit(X_o, y_o)

print(f"Eng yaxshi k: {qidiruv.best_params_}")
print(f"CV aniqlik:   {qidiruv.best_score_:.3f}")

KNeighborsClassifier(metric="euclidean")    # standart
KNeighborsClassifier(metric="manhattan")
KNeighborsClassifier(metric="minkowski", p=3)
KNeighborsClassifier(metric="cosine")

# Barcha qo'shnilar teng ovozga ega
KNeighborsClassifier(n_neighbors=5, weights="uniform")

# Yaqinroq qo'shni kuchliroq ovoz beradi
KNeighborsClassifier(n_neighbors=5, weights="distance")

from sklearn.neighbors import KNeighborsRegressor
from sklearn.metrics import mean_absolute_error

model = make_pipeline(
    StandardScaler(),
    KNeighborsRegressor(n_neighbors=5, weights="distance"),
)
model.fit(X_o, y_o)

bashorat = model.predict(X_s)
print(f"MAE: {mean_absolute_error(y_s, bashorat):.2f}")

KNeighborsClassifier(algorithm="ball_tree")   # ko'p o'lchamda yaxshi
KNeighborsClassifier(algorithm="kd_tree")     # kam o'lchamda tez
KNeighborsClassifier(algorithm="brute")       # to'g'ridan-to'g'ri
KNeighborsClassifier(algorithm="auto")        # o'zi tanlaydi
KNeighborsClassifier(n_jobs=-1)               # barcha yadrolar

from sklearn.neighbors import NearestNeighbors

# Har bir foydalanuvchining film baholari
baholar = np.array([
    [5, 4, 0, 0, 3],     # Husanboy
    [4, 5, 0, 1, 2],     # Jasur
    [0, 0, 5, 4, 0],     # Malika
    [1, 0, 4, 5, 1],     # Aziz
    [5, 5, 1, 0, 4],     # Nodira
])

izlovchi = NearestNeighbors(n_neighbors=3, metric="cosine")
izlovchi.fit(baholar)

masofalar, indekslar = izlovchi.kneighbors([baholar[0]])

ismlar = ["Husanboy", "Jasur", "Malika", "Aziz", "Nodira"]
print("Husanboyga o'xshash foydalanuvchilar:")
for masofa, idx in zip(masofalar[0][1:], indekslar[0][1:]):
    print(f"  {ismlar[idx]:10s} o'xshashlik: {1 - masofa:.3f}")

izlovchi = NearestNeighbors(n_neighbors=5).fit(X_oqitish)
masofalar, _ = izlovchi.kneighbors(X_sinov)

ortacha_masofa = masofalar.mean(axis=1)
chegara = np.percentile(ortacha_masofa, 95)

anomaliyalar = X_sinov[ortacha_masofa > chegara]
print(f"Topilgan anomaliyalar: {len(anomaliyalar)}")