Resonance-ML

A lightweight, simple Machine Learning classifier based on resonance frequency + cosine similarity. This package allows developers to train models using a CSV dataset and predict outcomes based on input features.

Features

• Train your ML model using a CSV file.
• Select columns for training and output class.
• Predict using user input.
• Extremely fast and lightweight.
• Works in React, Node.js, Express, and browser.js apps.

Installation

npm install resonance-ml

Project Structure

resonance-ml/
│
├── package.json
├── index.js
├── README.md
├── src/
│   └── ResonanceFrequencyClassifier.js

Quick Example (Node.js)

import { ResonanceFrequencyClassifier } from "resonance-ml";

const classifier = new ResonanceFrequencyClassifier();

// Training data
const trainingData = [
  { height: 170, weight: 65, gender: "male" },
  { height: 160, weight: 55, gender: "female" }
];

classifier.train(trainingData, ["height", "weight"], "gender");

// Prediction
const result = classifier.predict({ height: 168, weight: 62 });
console.log(result);

Usage in React + CSV Upload

Example UI flow:

1. User uploads CSV file
2. User selects training columns
3. User selects output column (class)
4. User inputs test values
5. Prediction displayed

Example React Component

import React, { useState } from "react";
import Papa from "papaparse";
import { ResonanceFrequencyClassifier } from "resonance-ml";

export default function MLTrainer() {
  const [csvData, setCsvData] = useState([]);
  const [columns, setColumns] = useState([]);
  const [features, setFeatures] = useState([]);
  const [target, setTarget] = useState("");
  const [prediction, setPrediction] = useState(null);

  const classifier = new ResonanceFrequencyClassifier();

  const handleCSV = (e) => {
    Papa.parse(e.target.files[0], {
      header: true,
      complete: (results) => {
        setCsvData(results.data);
        setColumns(Object.keys(results.data[0]));
      }
    });
  };

  const trainModel = () => {
    classifier.train(csvData, features, target);
  };

  const testPredict = (testInput) => {
    const result = classifier.predict(testInput);
    setPrediction(result);
  };

  return (
    <div>
      <h2>Resonance ML Trainer</h2>

      <input type="file" accept=".csv" onChange={handleCSV} />

      <h3>Select feature columns</h3>
      {columns.map((col) => (
        <label key={col}>
          <input
            type="checkbox"
            value={col}
            onChange={(e) =>
              setFeatures((prev) => [...prev, e.target.value])
            }
          />
          {col}
        </label>
      ))}

      <h3>Select target column</h3>
      <select onChange={(e) => setTarget(e.target.value)}>
        <option>Select...</option>
        {columns.map((col) => (
          <option key={col}>{col}</option>
        ))}
      </select>

      <button onClick={trainModel}>Train Model</button>

      <button
        onClick={() => testPredict({ height: 168, weight: 62 })}
      >
        Predict Test Input
      </button>

      {prediction && <h3>Prediction: {prediction}</h3>}
    </div>
  );
}

API Reference

train(data, featureColumns, targetColumn)

Trains the model.

• data: Array of objects (CSV rows)
• featureColumns: Columns used for training
• targetColumn: Column to predict

predict(inputObject)

Returns the predicted class based on similarity.

License

MIT

Author

Elanesan Kumaravel