@tensorscript/ts-deeplearning

Deep Learning Classification and Regression with Tensorflow (Clustering coming soon)

Full Documentation

Installation

$ npm i @tensorscript/ts-deeplearning

Usage

Classification

Test against the Iris Flower Data Set

import { DeepLearningClassification, } from '@tensorscript/ts-deeplearning';
import ms from 'modelscript';

async function main(){
  const irisFlowerDataCSV = await ms.csv.loadCSV('./test/mock/data/iris_data.csv');
  const DataSet = new ms.DataSet(irisFlowerDataCSV);
    /**
     * encodedData = [ 
     *  { sepal_length_cm: 5.1,
         sepal_width_cm: 3.5,
        petal_length_cm: 1.4,
        petal_width_cm: 0.2,
        plant: 'Iris-setosa',
        'plant_Iris-setosa': 1,
        'plant_Iris-versicolor': 0,
        'plant_Iris-virginica': 0 },
        ...
        { sepal_length_cm: 5.9,
        sepal_width_cm: 3,
        petal_length_cm: 4.2,
        petal_width_cm: 1.5,
        plant: 'Iris-versicolor',
        'plant_Iris-setosa': 0,
        'plant_Iris-versicolor': 1,
        'plant_Iris-virginica': 0 },
      ];
    */
  const encodedData = DataSet.fitColumns({
    columns: [
      {
        name: 'plant',
        options: {
          strategy: 'onehot',
        },
      },
    ],
    returnData:true,
  });
  const independentVariables = [
    'sepal_length_cm',
    'sepal_width_cm',
    'petal_length_cm',
    'petal_width_cm',
  ];
  const dependentVariables = [
    'plant_Iris-setosa',
    'plant_Iris-versicolor',
    'plant_Iris-virginica',
  ];
  const x_matrix = DataSet.columnMatrix(independentVariables); 
  const y_matrix = DataSet.columnMatrix(dependentVariables);
  /*
    x_matrix = [
      [ 5.1, 3.5, 1.4, 0.2 ],
      [ 4.9, 3, 1.4, 0.2 ],
      [ 4.7, 3.2, 1.3, 0.2 ],
      ...
    ]; 
    y_matrix = [
      [ 1, 0, 0 ],
      [ 1, 0, 0 ],
      [ 1, 0, 0 ],
      ...
    ] 
    */
  const input_x = [
    [5.1, 3.5, 1.4, 0.2, ],
    [6.3, 3.3, 6.0, 2.5, ],
    [5.6, 3.0, 4.5, 1.5, ],
    [5.0, 3.2, 1.2, 0.2, ],
    [4.5, 2.3, 1.3, 0.3, ],
  ];
  const nnClassification = new DeepLearningClassification();
  const nnClassificationModel = await nnClassification.train(x_matrix, y_matrix);
  const predictions = await nnClassification.predict(input_x);
  const answers = await nnClassification.predict(input_x, {
    probability:false,
  });
  /*
    predictions = [
      [ 0.989512026309967, 0.010471616871654987, 0.00001649192017794121, ],
      [ 0.0000016141033256644732, 0.054614484310150146, 0.9453839063644409, ],
      [ 0.001930746017023921, 0.6456733345985413, 0.3523959517478943, ],
      [ 0.9875779747962952, 0.01239941269159317, 0.00002274810685776174, ],
      [ 0.9545140862464905, 0.04520365223288536, 0.0002823179238475859, ],
    ];
    answers = [
      [ 1, 0, 0, ],
      [ 0, 0, 1, ],
      [ 0, 1, 0, ],
      [ 1, 0, 0, ],
      [ 1, 0, 0, ],
    ];
   */
}

main();

Regression

Test against the Boston Housing Data Set

import { DeepLearningRegression, } from '@tensorscript/ts-deeplearning';
import ms from 'modelscript';

function scaleColumnMap(columnName) {
  return {
    name: columnName,
    options: {
      strategy: 'scale',
      scaleOptions: {
        strategy:'standard'
      }
    }
  }
}

async function main(){
  const housingdataCSV = await ms.csv.loadCSV('./test/mock/data/boston_housing_data.csv');
  /*
  housingdataCSV = [
    { CRIM: 0.00632, ZN: 18, INDUS: 2.31, CHAS: 0, NOX: 0.538, RM: 6.575, AGE: 65.2, DIS: 4.09, RAD: 1, TAX: 296, PTRATIO: 15.3, B: 396.9, LSTAT: 4.98, MEDV: 24 },
    { CRIM: 0.02731, ZN: 0, INDUS: 7.07, CHAS: 0, NOX: 0.469, RM: 6.421, AGE: 78.9, DIS: 4.9671, RAD: 2, TAX: 242, PTRATIO: 17.8, B: 396.9, LSTAT: 9.14, MEDV: 21.6 },
    ...
  ]
  */
  const DataSet = new ms.DataSet(housingdataCSV);
  const independentVariables = [
    'CRIM',
    'ZN',
    'INDUS',
    'CHAS',
    'NOX',
    'RM',
    'AGE',
    'DIS',
    'RAD',
    'TAX',
    'PTRATIO',
    'B',
    'LSTAT',
  ];
  const dependentVariables = [
    'MEDV',
  ];
  const columns = independentVariables.concat(dependentVariables);
  DataSet.fitColumns({
    columns: columns.map(scaleColumnMap),
    returnData:false,
  });
  const x_matrix = DataSet.columnMatrix(independentVariables);
  const y_matrix = DataSet.columnMatrix(dependentVariables);
  /* x_matrix = [
    [ -0.41936692921321594, 0.2845482693404666, -1.2866362317172035, -0.272329067679207, -0.1440748547324509, 0.4132629204530747, -0.119894767215809, 0.1400749839795629, -0.981871187861867, -0.6659491794887338, -1.457557967289609, 0.4406158949991029, -1.074498970343932 ],
    [ -0.41692666996409716, -0.4872401872268264, -0.5927943782429392, -0.272329067679207, -0.7395303607434242, 0.1940823874370036, 0.3668034264326209, 0.5566090495704026, -0.8670244885881488, -0.9863533804386945, -0.3027944997494681, 0.4406158949991029, -0.49195252491856634 ]
    ...
  ];
  y_matrix = [
    [ 0.15952778852449556 ],
    [ -0.1014239172731213 ],
    ...
  ];
  const y_vector = ms.util.pivotVector(y_matrix)[ 0 ];// not used but just illustrative
  y_vector = [ 0.15952778852449556, -0.1014239172731213, ... ]
    */
  const input_x = [
    [-0.41936692921321594, 0.2845482693404666, -1.2866362317172035, -0.272329067679207, -0.1440748547324509, 0.4132629204530747, -0.119894767215809, 0.1400749839795629, -0.981871187861867, -0.6659491794887338, -1.457557967289609, 0.4406158949991029, -1.074498970343932,],
    [-0.41692666996409716, -0.4872401872268264, -0.5927943782429392, -0.272329067679207, -0.7395303607434242, 0.1940823874370036, 0.3668034264326209, 0.5566090495704026, -0.8670244885881488, -0.9863533804386945, -0.3027944997494681, 0.4406158949991029, -0.49195252491856634,],
  ];
  const nnRegression = new DeepLearningRegression();
  const model = await nnRegression.train(x_matrix, y_matrix);
  const predictions = await nnRegressionWide.predict(input_x); // [ [ 0.43396109342575073 ], [ 0.12437985092401505 ] ]
  const predictions_unscaled = predictions.map(pred=>DataSet.scalers.get('MEDV').descale(pred[0])); //[ 26.523991670220486, 23.67674075943165 ]
}

main();

Multiple Linear Regression

Test against the Portland housing price dataset

import { MultipleLinearRegression, } from '@tensorscript/ts-deeplearning';
import ms from 'modelscript';

function scaleColumnMap(columnName) {
  return {
    name: columnName,
    options: {
      strategy: 'scale',
      scaleOptions: {
        strategy:'standard'
      }
    }
  }
}

async function main(){
  const housingdataCSV = await ms.csv.loadCSV('./test/mock/data/portland_housing_data.csv');
  /*
  housingdataCSV = [
    { sqft: 2104, bedrooms: 3, price: 399900 },
    { sqft: 1600, bedrooms: 3, price: 329900 },
    ...
    { sqft: 1203, bedrooms: 3, price: 239500 }
  ]
  */
  const DataSet = new ms.DataSet(housingdataCSV);
  DataSet.fitColumns({
    columns: [
      'sqft',
      'bedrooms',
      'price',
    ].map(scaleColumnMap),
    returnData:true,
  });
  const independentVariables = [ 'sqft', 'bedrooms',];
  const dependentVariables = [ 'price', ];
  const x_matrix = DataSet.columnMatrix(independentVariables);
  const y_matrix = DataSet.columnMatrix(dependentVariables);
  /* x_matrix = [
      [2014, 3],
      [1600, 3],
    ];
    y_matrix = [
      [399900],
      [329900],
    ];
    const y_vector = ms.util.pivotVector(y_matrix)[ 0 ];// not used but just illustrative
    // y_vector = [ 399900, 329900]
   */
  const testSqft = DataSet.scalers.get('sqft').scale(1650);
  const testBedrooms = DataSet.scalers.get('bedrooms').scale(3);
  const input_x = [
    testSqft,
    testBedrooms,
  ]; // input_x: [ -0.4412732005944351, -0.2236751871685913 ]
  const tfMLR = new MultipleLinearRegression();
  const model = await tfMLR.train(x_matrix, y_matrix);
  const scaledPrediction = await tfMLR.predict(input_x); // [ -0.3785287367962629 ]
  const prediction = DataSet.scalers.get('price').descale(scaledPrediction); // prediction: 293081.4643348962
}

main();

Logistic Regression

Test against the Social Media Ads

import { LogisticRegression, } from '@tensorscript/ts-deeplearning';
import ms from 'modelscript';

function scaleColumnMap(columnName) {
  return {
    name: columnName,
    options: {
      strategy: 'scale',
      scaleOptions: {
        strategy:'standard'
      }
    }
  }
}

async function main(){
  const CSVData = await ms.csv.loadCSV('./test/mock/data/social_network_ads.csv');
  const DataSet = new ms.DataSet(CSVData);
  const scaledData = DataSet.fitColumns({
    columns: independentVariables.map(scaleColumnMap),
    returnData:true,
  });
  /*
    scaledData = [
      { 'User ID': 15624510,
         Gender: 'Male',
         Age: -1.7795687879022388,
         EstimatedSalary: -1.4881825118632386,
         Purchased: 0 },
      { 'User ID': 15810944,
         Gender: 'Male',
         Age: -0.253270175924977,
         EstimatedSalary: -1.458854384319991,
         Purchased: 0 },
      ...
    ];
    */
  const independentVariables = [
    'Age',
    'EstimatedSalary',
  ];
  const dependentVariables = [
    'Purchased',
  ];
  const x_matrix = DataSet.columnMatrix(independentVariables);
  const y_matrix = DataSet.columnMatrix(dependentVariables);
  /*
    x_matrix = [
      [ -1.7795687879022388, -1.4881825118632386 ],
      [ -0.253270175924977, -1.458854384319991 ],
      ...
    ];
    y_matrix = [
      [ 0 ],
      [ 0 ],
      ...
    ];
    */
  const input_x = [
    [-0.062482849427819266, 0.30083326827486173,], //0
    [0.7960601198093905, -1.1069168538010206,], //1
    [0.7960601198093905, 0.12486450301537644,], //0
    [0.4144854668150751, -0.49102617539282206,], //0
    [0.3190918035664962, 0.5061301610775946,], //1
  ];
  const tfLR = new LogisticRegression();
  const model = await tfLR.train(x_matrix, y_matrix);
  const prediction = await tfLR.predict(input_x); // => [ [ 0 ], [ 0 ], [ 1 ], [ 0 ], [ 1 ] ],
}

main();

Testing

$ npm i
$ npm test

Contributing

Fork, write tests and create a pull request!

Misc

As of Node 8, ES modules are still used behind a flag, when running natively as an ES module

$ node --experimental-modules my-machine-learning-script.mjs
# Also there are native bindings that require Python 2.x, make sure if you're using Andaconda, you build with your Python 2.x bin
$ npm i --python=/usr/bin/python

License

MIT