mlbuddy-learn 0.1.3

An ML library that guides beginners step by step

🤖 MLBuddy

Your friendly machine learning companion. Build models, understand what they learned, and receive step-by-step guidance to improve them — all with just a few lines of code.

MLBuddy is designed to bridge the gap between complex machine learning workflows and beginners. Instead of treating models as a "black box," MLBuddy actively guides you through the process, explains your model's decisions visually, and suggests actionable next steps to boost performance.

---

Key Features

• Automated Model Comparison (mlbuddy.compare)**: Compare multiple ML architectures (Logistic Regression, Decision Trees, Random Forests, SVMs) in one command to find the absolute best match for your dataset.
• Transparent Explanations (model.explain())**: Instantly generate feature importance bar charts or coefficients to see exactly which features drive your model's predictions.
• Actionable Feedback suggestions (model.suggest())**: Receive direct, intelligent hints on how to improve your model based on its performance metrics (e.g., handling overfitting, engineering features, or tuning hyperparameters).
• Zero-Config Data Prep (mlbuddy.load_csv)**: Automatically handles data scaling and splitting with intelligent defaults.

---

Quickstart

Get up and running in less than 60 seconds.

1. Installation

pip install mlbuddy-learn

2. The Golden Workflow

Here is how simple it is to load, compare, train, explain, and improve a machine learning model:

import mlbuddy as ml

# 1. Load and prepare your dataset
X_train, X_test, y_train, y_test = ml.load_csv("your_data.csv", target_column="hired")

# 2. Compare all models to find the winner
results = ml.compare(X_train, X_test, y_train, y_test)

# 3. Train the best performing model
model = ml.train(X_train, X_test, y_train, y_test, model="random_forest")

# 4. Peer into the black box (generates a feature importance plot!)
model.explain()

# 5. Get tailored recommendations to improve accuracy
model.suggest()

---

Project Architecture

mlbuddy/
├── auto/
│   ├── data.py         # Scaling, train/test splitting, and CSV preprocessing
│   └── trainer.py      # Guided Model wrapper, training routines, and model comparison
├── explain/
│   └── visualizer.py   # Automatic feature importance & coefficient plotting
└── guide/
    └── suggest.py      # Rule-based suggestions engine tailored to accuracy thresholds

---

How it Guides You

When you call model.suggest(), MLBuddy analyzes your model's accuracy and provides contextual advice:

• Low Accuracy (< 60%): Recommends looking into data quality, checking labels, or gathering more samples.
• Moderate Accuracy (60% - 80%): Suggests trying more robust architectures (like Random Forests) and hyperparameter tuning.
• High Accuracy (> 90%): Flags potential overfitting risks, duplicates in the dataset, and guides you on preparing for production.

---

Development Setup

Clone the repository and set up a local development environment:

# Clone the repository
git clone https://github.com/Mohammedjaasir/mlbuddy-learn.git
cd mlbuddy-learn

# Install in editable mode with dev dependencies
pip install -e ".[dev]"

# Run tests
pytest

---

📄 License

Distributed under the MIT License. See LICENSE for more information.