يعني إيه Machine Learning؟

import numpy as np
import pandas as pd

rng = np.random.default_rng(42)

# ---------- apartments.csv (regression) ----------
n = 600
city = rng.choice(["Cairo", "Giza", "Alexandria"], n, p=[0.45, 0.30, 0.25])
area = rng.integers(60, 221, n)
rooms = np.clip((area / 45).round().astype(int) + rng.integers(-1, 2, n), 1, 6)
floor = rng.integers(0, 16, n)
age = rng.integers(0, 41, n)
metro_km = rng.uniform(0.2, 8, n).round(1)
city_bonus = pd.Series(city).map({"Cairo": 600, "Giza": 300, "Alexandria": 0}).to_numpy()
price = (area * 18 + rooms * 80 + floor * 15 - age * 12 - metro_km * 90
         + city_bonus + rng.normal(0, 250, n))
apartments = pd.DataFrame({
    "city": city, "area": area, "rooms": rooms, "floor": floor,
    "age": age, "metro_km": metro_km, "price_k": price.round(0).astype(int),
})
apartments.to_csv("apartments.csv", index=False)

# ---------- churn.csv (classification) ----------
n = 1500
tenure = rng.integers(1, 73, n)                     # months with the company
bill = rng.uniform(100, 900, n).round(0)            # monthly bill (EGP)
complaints = rng.poisson(0.8, n)
data_gb = rng.uniform(1, 80, n).round(1)
contract = rng.choice(["monthly", "yearly"], n, p=[0.6, 0.4])
logit = (-0.3 - 0.05 * tenure + 0.003 * (bill - 400) + 0.8 * complaints
         - 1.6 * (contract == "yearly") + 0.9 * ((data_gb > 50) & (tenure < 12)))
churned = (rng.random(n) < 1 / (1 + np.exp(-logit))).astype(int)
churn = pd.DataFrame({
    "tenure": tenure, "bill": bill, "complaints": complaints,
    "data_gb": data_gb, "contract": contract, "churned": churned,
})
churn.to_csv("churn.csv", index=False)

print("apartments.csv:", apartments.shape)
print("churn.csv:", churn.shape, "| churn rate:", f"{churned.mean():.0%}")

import pandas as pd
from sklearn.model_selection import train_test_split

df = pd.read_csv("apartments.csv")

X = df[["area", "rooms", "floor", "age", "metro_km"]]   # features
y = df["price_k"]                                        # target

X_train, X_test, y_train, y_test = train_test_split(
    X, y, test_size=0.2, random_state=42)   # 80% train, 20% test

print("Train:", X_train.shape)
print("Test: ", X_test.shape)