mini-scikit-learn 0.1.4

A brief description of your package

mini_scikit_learn

mini_scikit_learn is a minimalistic clone of scikit-learn, designed to provide essential machine learning functionalities with minimal overhead. It only depends on numpy and offers a variety of basic and advanced models, utility functions, metrics, data transformers, and model selection tools.

Features

Models

• Linear Models: Implementations of linear regression, logistic regression, etc.
• Tree-Based Models: Decision trees for classification and regression.
• K-Nearest Neighbors: KNN for classification and regression.
• Random Forest: Ensemble method for classification and regression.
• SVM: Support Vector Machines for classification and regression.(Under Testing)
• Naive Bayes: Various naive Bayes classifiers.
• Neural Networks: Basic feedforward and backpropagation neural networks, with customizable layers and activation functions.
• Ensembling Techniques: Advanced techniques such as voting, stacking, and boosting (AdaBoost, Gradient Boosting).

Utility Functions

• Cross Validation: Functions for k-fold cross-validation, train-test splitting, etc.
• Train-Test Split: Simple utility to split datasets into training and testing sets.
• K-Folds: Functionality to split data into k folds for cross-validation.

Metrics

• Accuracy: Measure the accuracy of predictions.
• Precision: Calculate the precision for classification models.
• Recall: Compute the recall for classification models.
• F1 Score: Calculate the F1 score for classification models.
• Mean Squared Error (MSE): Compute the mean squared error for regression models.
• Mean Absolute Error (MAE): Compute the mean absolute error for regression models.
• Log Loss: Calculate the logistic loss for classification models.

Data Transformers

• Encoders: Various encoding techniques for categorical data.
• Imputers: Different strategies for handling missing values, SimpleImputer, IterativeImputer, KNNImputer.
• Scalers: MinMaxScaler and StandardScaler for feature scaling.

Model Selection

• Grid Search: Exhaustive search over specified parameter values for an estimator.
• Random Search: Random search over specified parameter values for an estimator.

System Architecture

The design of mini_scikit_learn is heavily inspired by scikit-learn, with a strong use of inheritance from abstract classes such as Estimator and Predictor. The library respects the fit, predict, and transform API for all models and transformers, ensuring consistency and ease of use.

Core Components

• Estimator: Base class for all estimators in the library. Defines the fit method.
• Predictor: Base class for all predictors, extending Estimator with the predict method.
• Transformer: Base class for all transformers, extending Estimator with the transform method.

By adhering to these interfaces, mini_scikit_learn ensures that all components can be used interchangeably, promoting modularity and ease of integration.

Installation

To install mini_scikit_learn, you can use pip:

pip install mini_scikit_learn

Requirements

• Python > 3
• numpy

Example Usage

from mini_scikit_learn.model_selection import GridSearch
from mini_scikit_learn.linear_model import LinearRegression
from mini_scikit_learn.metrics import accuracy_score
from mini_scikit_learn.datasets import load_iris
from mini_scikit_learn.utils import train_test_split

# Load dataset
iris = load_iris()
X, y = iris.data, iris.target

# Split the data
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.25, random_state=42)

# Define the model
model = LinearRegression()

# Define parameter grid
param_grid = {'alpha': [0.1, 0.01, 0.001]}

# Perform grid search
grid_search = GridSearch(model, param_grid)
grid_search.fit(X_train, y_train)

# Get the best model
best_model = grid_search.get_best_params()
print("Best Parameters:", best_model)

# Evaluate the model
accuracy = accuracy_score(y_test, best_model.predict(X_test))
print(f"Accuracy: {accuracy}")

Documentation

For more detailed documentation and examples, please refer to the official mini_scikit_learn documentation.

Contributing

Contributions are welcome! Please fork the repository and submit pull requests.

License

This project is licensed under the MIT License. See the LICENSE file for details.

Acknowledgements

Inspired by the simplicity and efficiency of scikit-learn. This project aims to provide a lightweight alternative for quick prototyping and educational purposes.