dolphinboost 0.1.0

A lightweight machine learning library with educational implementations built from scratch.

DolphinBoost

DolphinBoost is a lightweight machine learning library built from scratch in Python for learning and understanding core machine learning algorithms.

The goal of DolphinBoost is to provide clean, readable implementations that help students and developers understand how machine learning models work internally.

Features

• Simple Linear Regression
• NumPy-based implementation
• Minimal dependencies
• Educational and beginner-friendly
• Easy-to-read source code

Installation

pip install dolphinboost

Quick Start

from dolphinboost import SimpleLinearRegression

X = [1, 2, 3, 4, 5]
y = [2, 4, 5, 4, 5]

model = SimpleLinearRegression()
model.fit(X, y)

print("Coefficient:", model.coef_)
print("Intercept:", model.intercept_)

predictions = model.predict([6, 7, 8])
print(predictions)

Example

import pandas as pd
from sklearn.model_selection import train_test_split
from dolphinboost import SimpleLinearRegression

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

X = df["YearsExperience"]
y = df["Salary"]

X_train, X_test, y_train, y_test = train_test_split(
    X, y, test_size=0.2, random_state=42
)

model = SimpleLinearRegression()
model.fit(X_train, y_train)

predictions = model.predict(X_test)

print(predictions)

API

SimpleLinearRegression

fit(X, y)

Train the model using the least-squares method.

predict(X)

Predict target values.

Attributes

Attribute	Description
coef_	Slope of the regression line
intercept_	Intercept of the regression line
beta_	Full parameter vector
X_mean_	Mean of input values
y_mean_	Mean of target values

Mathematical Foundation

Slope:

β₁ = Σ[(x - x̄)(y - ȳ)] / Σ[(x - x̄)²]

Intercept:

β₀ = ȳ − β₁x̄

Prediction:

y = β₀ + β₁x

Roadmap

Upcoming algorithms:

• Multiple Linear Regression
• Logistic Regression
• Metrics (MSE, MAE, R²)
• And Others

Why DolphinBoost?

• Learn machine learning by reading code.
• Lightweight and easy to understand.
• Built for students and educational purposes.
• No hidden abstractions.

License

MIT License

Author

Peddakotla Karthikeya