mathdatech/Python
1
0
Fork 0
mirror of https://github.com/TheAlgorithms/Python.git synced 2024-11-23 21:11:08 +00:00
Code Issues Packages Projects Releases Wiki Activity

[pre-commit.ci] auto fixes from pre-commit.com hooks

Browse Source 
for more information, see https://pre-commit.ci
This commit is contained in:
pre-commit-ci[bot] 2024-10-16 03:49:29 +00:00
parent b91bc91ee5
commit e23173e6a3
1 changed files with 16 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

22
machine_learning/ridge_regression.py
View File

@ -1,6 +1,7 @@
import numpy as np import numpy as np
import requests import requests
def collect_dataset(): def collect_dataset():
"""Collect dataset of CSGO """Collect dataset of CSGO
The dataset contains ADR vs Rating of a Player The dataset contains ADR vs Rating of a Player
@ -20,6 +21,7 @@ def collect_dataset():
dataset = np.matrix(data) dataset = np.matrix(data)
return dataset return dataset
def run_steep_gradient_descent(data_x, data_y, len_data, alpha, theta, lambda_reg): def run_steep_gradient_descent(data_x, data_y, len_data, alpha, theta, lambda_reg):
"""Run steep gradient descent and updates the Feature vector accordingly """Run steep gradient descent and updates the Feature vector accordingly
:param data_x : contains the dataset :param data_x : contains the dataset
@ -36,7 +38,7 @@ def run_steep_gradient_descent(data_x, data_y, len_data, alpha, theta, lambda_re
prod = np.dot(theta, data_x.transpose()) prod = np.dot(theta, data_x.transpose())
prod -= data_y.transpose() prod -= data_y.transpose()
sum_grad = np.dot(prod, data_x) sum_grad = np.dot(prod, data_x)
# Add regularization to the gradient # Add regularization to the gradient
theta_regularized = np.copy(theta) theta_regularized = np.copy(theta)
theta_regularized[0, 0] = 0 # Don't regularize the bias term theta_regularized[0, 0] = 0 # Don't regularize the bias term
@ -45,6 +47,7 @@ def run_steep_gradient_descent(data_x, data_y, len_data, alpha, theta, lambda_re
theta = theta - (alpha / n) * sum_grad theta = theta - (alpha / n) * sum_grad
return theta return theta
def sum_of_square_error(data_x, data_y, len_data, theta, lambda_reg): def sum_of_square_error(data_x, data_y, len_data, theta, lambda_reg):
"""Return sum of square error for error calculation """Return sum of square error for error calculation
:param data_x : contains our dataset :param data_x : contains our dataset
@ -57,12 +60,15 @@ def sum_of_square_error(data_x, data_y, len_data, theta, lambda_reg):
prod = np.dot(theta, data_x.transpose()) prod = np.dot(theta, data_x.transpose())
prod -= data_y.transpose() prod -= data_y.transpose()
sum_elem = np.sum(np.square(prod)) sum_elem = np.sum(np.square(prod))
# Add regularization to the cost function # Add regularization to the cost function
regularization_term = lambda_reg * np.sum(np.square(theta[:, 1:])) # Don't regularize the bias term regularization_term = lambda_reg * np.sum(
np.square(theta[:, 1:])
) # Don't regularize the bias term
error = (sum_elem / (2 * len_data)) + (regularization_term / (2 * len_data)) error = (sum_elem / (2 * len_data)) + (regularization_term / (2 * len_data))
return error return error
def run_ridge_regression(data_x, data_y, lambda_reg=1.0): def run_ridge_regression(data_x, data_y, lambda_reg=1.0):
"""Implement Ridge Regression over the dataset """Implement Ridge Regression over the dataset
:param data_x : contains our dataset :param data_x : contains our dataset
@ -79,12 +85,15 @@ def run_ridge_regression(data_x, data_y, lambda_reg=1.0):
theta = np.zeros((1, no_features)) theta = np.zeros((1, no_features))
for i in range(iterations): for i in range(iterations):
theta = run_steep_gradient_descent(data_x, data_y, len_data, alpha, theta, lambda_reg) theta = run_steep_gradient_descent(
data_x, data_y, len_data, alpha, theta, lambda_reg
)
error = sum_of_square_error(data_x, data_y, len_data, theta, lambda_reg) error = sum_of_square_error(data_x, data_y, len_data, theta, lambda_reg)
print(f"At Iteration {i + 1} - Error is {error:.5f}") print(f"At Iteration {i + 1} - Error is {error:.5f}")
return theta return theta
def mean_absolute_error(predicted_y, original_y): def mean_absolute_error(predicted_y, original_y):
"""Return mean absolute error for error calculation """Return mean absolute error for error calculation
:param predicted_y : contains the output of prediction (result vector) :param predicted_y : contains the output of prediction (result vector)
@ -94,6 +103,7 @@ def mean_absolute_error(predicted_y, original_y):
total = sum(abs(y - predicted_y[i]) for i, y in enumerate(original_y)) total = sum(abs(y - predicted_y[i]) for i, y in enumerate(original_y))
return total / len(original_y) return total / len(original_y)
def main(): def main():
"""Driver function""" """Driver function"""
data = collect_dataset() data = collect_dataset()
@ -104,12 +114,12 @@ def main():
lambda_reg = 1.0 # Set your desired regularization parameter lambda_reg = 1.0 # Set your desired regularization parameter
theta = run_ridge_regression(data_x, data_y, lambda_reg) theta = run_ridge_regression(data_x, data_y, lambda_reg)
len_result = theta.shape[1] len_result = theta.shape[1]
print("Resultant Feature vector : ") print("Resultant Feature vector : ")
for i in range(len_result): for i in range(len_result):
print(f"{theta[0, i]:.5f}") print(f"{theta[0, i]:.5f}")
if __name__ == "__main__": if __name__ == "__main__":
main() main()