Logistic Regression

In logistic regression model, we assume a linear relationship between predictor varaibles and log-odds of the event that \(Y = 1\).

\[l = \log \frac{p}{1-p} = \beta_0 + \beta_1 x_1 + \beta_2 x_2 + ....\]

The objective of logistic model is to find \(\beta_0, \beta_1, \beta_2...\) above to best fit this assumption.

In datacook, we implement logistic regression by stochatic gradient descent (SGD).

Import

import * as Datacook from 'datacook';
const { LogisticRegression } = DataCook.Model;

const lr = new LogisticRegression({});

parameter	type	description
penalty	‘l1’ \| ‘l2’ \| ‘none’	penalty used in the penalization. default = ‘none’
fitIntercept	boolean	Whether to calculate the intercept for this model. If set to False, no intercept will be used in calculations. default = false
c	number	Regularization strength; must be a positive float. Larger values specify stronger regularization. default = 1.
optimizerType	‘sgd’ \| ‘momentum’ \| ‘adagrad’ \| ‘adadelta’ \| ‘adam’ \| ‘adamax’ \| ‘rmsprop’	optimizer types for training. All of the following optimizers types supported in tensorflow.js can be applied. Default to ‘adam’
optimizerProps	OptimizerProps	parameters used to init corresponding optimizer, you can refer to documentations in tensorflow.js to find the supported initailization paratemters for a given type of optimizer. For example, `{ learningRate: 0.1, beta1: 0.1, beta2: 0.2, epsilon: 0.1 }` could be used to initialize adam optimizer.

Fit logistic regression model according to X, y.

async fit(xData: Tensor | RecursiveArray<number>,
    yData: Tensor | RecursiveArray<number>,
    params?: LogisticRegressionTrainParams

parameter	type	description
xData	Tensor \| RecursiveArray	Tensor like of shape (n_samples, n_features), input feature
yData	Tensor \| RecursiveArray	Tensor like of shape (n_sample, ), input target values
params	LogisticRegressionTrainParams	training parameters, batchSize: batch size: default to 32, maxIterTimes: max iteration times, default to 20000

Make predictions using logistic regression model.

async predict(xData: Tensor | RecursiveArray<number>): Promise<Tensor>

parameter	type	description
xData	Tensor	RecursiveArray	Input features

Predicted classes

Training logistic regression model by batch

parameter	type	description
xData	Tensor \| RecursiveArray	Tensor like of shape (n_samples, n_features), input feature
yData	Tensor \| RecursiveArray	Tensor like of shape (n_sample, ), input target values

LogisticRegression

Predict probabilities using logistic regression model.

async predictProba(xData: Tensor | RecursiveArray<number>): Promise<Tensor>

parameter	type	description
xData	Tensor \| RecursiveArray	Input features

Predicted probabilities

Get coefficients of logistic model

getCoef(): { coefficients: Tensor, intercept: Tensor }

{ coefficients: Tensor, intercept: Tensor }

Load model paramters from json string object

async fromJson(modelJson: string)

parameter	type	description
modelJson	string	model json string

Dump model parameters to json string.

async toJson(): Promise<string>

string of model json