67 lines
2.0 KiB
Python
67 lines
2.0 KiB
Python
from sklearn.linear_model import Lasso
|
|
from sklearn.ensemble import RandomForestRegressor
|
|
from sklearn.feature_selection import f_regression
|
|
from sklearn.preprocessing import MinMaxScaler
|
|
import numpy as np
|
|
|
|
np.random.seed(0)
|
|
|
|
# Генерация данных
|
|
size = 500
|
|
X = np.random.uniform(0, 1, (size, 15))
|
|
Y = (10 * np.sin(np.pi * X[:, 0] * X[:, 1]) + 20 * (X[:, 2] - .5) ** 2 +
|
|
10 * X[:, 3] + 5 * X[:, 4] ** 5 + np.random.normal(0, 1))
|
|
X[:, 10:] = X[:, :5] + np.random.normal(0, .025, (size, 5))
|
|
|
|
# Имена признаков
|
|
names = ["x%s" % i for i in range(1, 16)]
|
|
|
|
# Ранги признаков
|
|
ranks = {}
|
|
|
|
|
|
# Функция для расчета рангов
|
|
def calculate_ranks(method, X, Y):
|
|
if method == "Lasso":
|
|
model = Lasso(alpha=0.5)
|
|
elif method == "Random Forest":
|
|
model = RandomForestRegressor(n_estimators=100)
|
|
elif method == "f_regression":
|
|
f_scores, _ = f_regression(X, Y)
|
|
return dict(zip(names, f_scores))
|
|
model.fit(X, Y)
|
|
return dict(zip(names, model.coef_ if method == "Lasso" else model.feature_importances_))
|
|
|
|
|
|
# Ранг для каждого метода
|
|
for method in ["Lasso", "Random Forest", "f_regression"]:
|
|
ranks[method] = calculate_ranks(method, X, Y)
|
|
|
|
|
|
# Нормализация рангов
|
|
def create_normalized_rank_dict(ranks, names):
|
|
ranks = np.abs(ranks)
|
|
minmax = MinMaxScaler()
|
|
ranks = minmax.fit_transform(
|
|
np.array(ranks).reshape(15, 1)).ravel()
|
|
ranks = map(lambda x: round(x, 2), ranks)
|
|
return dict(zip(names, ranks))
|
|
|
|
|
|
# Среднее значение рангов
|
|
mean = {}
|
|
for key, value in ranks.items():
|
|
for item in value.items():
|
|
if (item[0] not in mean):
|
|
mean[item[0]] = 0
|
|
mean[item[0]] += item[1]
|
|
|
|
# Сортируем признаки по среднему значению рангов
|
|
sorted_mean = sorted(mean.items(), key=lambda x: x[1], reverse=True)
|
|
|
|
# Вывод признаков и их рангов
|
|
result = {}
|
|
for item in sorted_mean:
|
|
result[item[0]] = item[1]
|
|
print(f'{item[0]}: {item[1]}')
|