External module "ml/incr/binary-classification/docs/types/index.d"

Export assignment incrBinaryClassification

• incrBinaryClassification(N: number, options?: Options): Accumulator

• Returns an accumulator function which incrementally performs binary classification using stochastic gradient descent (SGD).

  Method

  • The sub-gradient of the loss function is estimated for each datum and the classification model is updated incrementally, with a decreasing learning rate and regularization of model feature weights using L2 regularization.
  • Stochastic gradient descent is sensitive to the scaling of the features. One is advised to either scale each feature to [0,1] or [-1,1] or to transform each feature into z-scores with zero mean and unit variance. One should keep in mind that the same scaling has to be applied to training data in order to obtain accurate predictions.
  • In general, the more data provided to an accumulator, the more reliable the model predictions.

  References

  • Shalev-Shwartz, S., Singer, Y., Srebro, N., & Cotter, A. (2011). Pegasos: Primal estimated sub-gradient solver for SVM. Mathematical Programming, 127(1), 3–30. doi:10.1007/s10107-010-0420-4

  throws

  first argument must be a positive integer

  throws

  must provide valid options

  Parameters

  • N: number

    number of features

  • Optional options: Options

    options object

  Returns Accumulator

  accumulator function

  Example

  var Float64Array = require( `@stdlib/array/float64` );
  var array = require( `@stdlib/ndarray/array` );
  
  // Create an accumulator:
  var accumulator = incrBinaryClassification( 3, {
      'intercept': true,
      'lambda': 1.0e-5
  });
  
  // ...
  
  // Update the model:
  var x = array( new Float64Array( [ 2.3, 1.0, 5.0 ] ) );
  var coefs = accumulator( x, 1 );
  // returns <ndarray>
  
  // ...
  
  // Create a new observation vector:
  x = array( new Float64Array( [ 2.3, 5.3, 8.6 ] ) );
  
  // Predict the response value:
  var yhat = accumulator.predict( x );
  // returns <ndarray>