noc_ml_dart 1.0.0

A Dart/Flutter version of NocML - a lightweight machine learning library featuring Classification (KNN, Naive Bayes, Logistic Regression), Clustering, and Forecasting.

NocML_Dart 🚀

[Also available in Indonesian / Tersedia dalam Bahasa Indonesia: README_id.md]

A Dart & Flutter version of the NocML library, developed by the Nocturnailed Community. This package brings lightweight, efficient machine learning capabilities directly to Dart without requiring complex integrations or heavy external C++ components.

🌟 Overview

noc_ml_dart is designed strictly as a port of our lightweight machine learning engine geared initially toward microcontrollers, translated idiomatically to Dart for pure cross-platform development (Web, iOS, Android, Desktop).

Features:

• Preprocessing: StandardScaler, MinMaxScaler
• Classification: K-Nearest Neighbors (KNN), Naive Bayes, Logistic Regression
• Clustering: K-Means
• Forecasting / Regression: Simple Linear Regression
• Evaluation Metrics: Accuracy, MSE, MAE, R2-Score
• Zero Dependencies: Pure Dart implementation

📦 Installation

Add the following to your pubspec.yaml file:

dependencies:
  noc_ml_dart: ^1.0.0

Then run dart pub get or flutter pub get.

🚀 Quick Start

Here are a few examples showcasing noc_ml_dart.

1. Preprocessing Data

import 'package:noc_ml_dart/noc_ml_dart.dart';

void main() {
  List<List<double>> rawData = [
    [10.0, 50.0],
    [20.0, 30.0],
    [30.0, 40.0]
  ];

  var scaler = StandardScaler();
  var scaledData = scaler.fitTransform(rawData);
  print(scaledData);
}

2. K-Nearest Neighbors (KNN) & Evaluation

import 'package:noc_ml_dart/noc_ml_dart.dart';

void main() {
  var knn = KNN(k: 3);
  
  // Training Data: [feature1, feature2]
  List<List<double>> X_train = [[1.0, 2.0], [1.5, 1.8], [5.0, 8.0], [8.0, 8.0]];
  // Labels
  List<int> y_train = [0, 0, 1, 1];
  
  knn.fit(X_train, y_train);
  
  // Predict
  List<int> y_pred = [knn.predict([1.1, 1.9]), knn.predict([6.0, 9.0])];
  List<int> y_true = [0, 1];
  
  // Evaluate
  double acc = EvaluationMetrics.accuracyScore(y_true, y_pred);
  print('Accuracy: $acc'); // Output: 1.0 (100%)
}

3. Forecasting (Simple Linear Regression)

import 'package:noc_ml_dart/noc_ml_dart.dart';

void main() {
  var regression = SimpleLinearRegression();
  regression.fit([1, 2, 3, 4, 5], [2, 4, 6, 8, 10]);
  
  double forecast = regression.predict(6); 
  print('Forecast: $forecast'); // Output: 12.0
}

📚 Supported Algorithms & Functions

Preprocessing

• StandardScaler: Standardizes features by removing the mean and scaling to unit variance (Z-score).
• MinMaxScaler: Transforms features by scaling each feature to a given range (default 0 to 1).

Evaluation Metrics

• accuracyScore(): Ratio of correct predictions to total predictions.
• meanSquaredError(): Averages the squared differences between predicted and actual values.
• meanAbsoluteError(): Averages the absolute differences.
• r2Score(): Coefficient of determination indicating the proportion of the variance in the dependent variable.

Classification

• Logistic Regression: Iterative gradient-descent binary classification.
• Naive Bayes: Fast Gaussian probabilistic classifier.
• K-Nearest Neighbors (KNN): Distance-based classification.

Clustering

• K-Means: Simple and effective unsupervised learning algorithm for partitioning data into k clusters.

Forecasting & Regression

• Simple Linear Regression: Trend line prediction and standard linear modeling.

🤝 Contributing

Contributions, issues, and feature requests are welcome! Feel free to check issues page.

📝 License

This project is MIT licensed.