work/smartcore
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat: NB documentation

Browse Source
This commit is contained in:
Volodymyr Orlov
2020-12-17 19:00:11 -08:00
parent 2c892aa603
commit 5a185479a7
6 changed files with 167 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/naive_bayes/bernoulli.rs
+35
View File
@@ -1,3 +1,38 @@
//! # Bernoulli Naive Bayes
//!
//! Bernoulli Naive Bayes classifier is a variant of [Naive Bayes](../index.html) for the data that is distributed according to multivariate Bernoulli distribution.
//! It is used for discrete data with binary features. One example of a binary feature is a word that occurs in the text or not.
//!
//! Example:
//!
//! ```
//! use smartcore::linalg::naive::dense_matrix::*;
//! use smartcore::naive_bayes::bernoulli::BernoulliNB;
//!
//! // Training data points are:
//! // Chinese Beijing Chinese (class: China)
//! // Chinese Chinese Shanghai (class: China)
//! // Chinese Macao (class: China)
//! // Tokyo Japan Chinese (class: Japan)
//! let x = DenseMatrix::<f64>::from_2d_array(&[
//! &[1., 1., 0., 0., 0., 0.],
//! &[0., 1., 0., 0., 1., 0.],
//! &[0., 1., 0., 1., 0., 0.],
//! &[0., 1., 1., 0., 0., 1.],
//! ]);
//! let y = vec![0., 0., 0., 1.];
//!
//! let nb = BernoulliNB::fit(&x, &y, Default::default()).unwrap();
//!
//! // Testing data point is:
//! // Chinese Chinese Chinese Tokyo Japan
//! let x_test = DenseMatrix::<f64>::from_2d_array(&[&[0., 1., 1., 0., 0., 1.]]);
//! let y_hat = nb.predict(&x_test).unwrap();
//! ```
//!
//! ## References:
//!
//! * ["Introduction to Information Retrieval", Manning C. D., Raghavan P., Schutze H., 2009, Chapter 13 ](https://nlp.stanford.edu/IR-book/information-retrieval-book.html)
use crate::error::Failed;
use crate::linalg::row_iter;
use crate::linalg::BaseVector;
src/naive_bayes/categorical.rs
+32
View File
@@ -1,3 +1,35 @@
//! # Categorical Naive Bayes
//!
//! Categorical Naive Bayes is a variant of [Naive Bayes](../index.html) for the categorically distributed data.
//! It assumes that each feature has its own categorical distribution.
//!
//! Example:
//!
//! ```
//! use smartcore::linalg::naive::dense_matrix::*;
//! use smartcore::naive_bayes::categorical::CategoricalNB;
//!
//! let x = DenseMatrix::from_2d_array(&[
//! &[3., 4., 0., 1.],
//! &[3., 0., 0., 1.],
//! &[4., 4., 1., 2.],
//! &[4., 2., 4., 3.],
//! &[4., 2., 4., 2.],
//! &[4., 1., 1., 0.],
//! &[1., 1., 1., 1.],
//! &[0., 4., 1., 0.],
//! &[0., 3., 2., 1.],
//! &[0., 3., 1., 1.],
//! &[3., 4., 0., 1.],
//! &[3., 4., 2., 4.],
//! &[0., 3., 1., 2.],
//! &[0., 4., 1., 2.],
//! ]);
//! let y = vec![0., 0., 1., 1., 1., 0., 1., 0., 1., 1., 1., 1., 1., 0.];
//!
//! let nb = CategoricalNB::fit(&x, &y, Default::default()).unwrap();
//! let y_hat = nb.predict(&x).unwrap();
//! ```
use crate::error::Failed;
use crate::linalg::BaseVector;
use crate::linalg::Matrix;
src/naive_bayes/gaussian.rs
+24
View File
@@ -1,3 +1,27 @@
//! # Gaussian Naive Bayes
//!
//! Gaussian Naive Bayes is a variant of [Naive Bayes](../index.html) for the data that follows Gaussian distribution and
//! it supports continuous valued features conforming to a normal distribution.
//!
//! Example:
//!
//! ```
//! use smartcore::linalg::naive::dense_matrix::*;
//! use smartcore::naive_bayes::gaussian::GaussianNB;
//!
//! let x = DenseMatrix::from_2d_array(&[
//! &[-1., -1.],
//! &[-2., -1.],
//! &[-3., -2.],
//! &[ 1., 1.],
//! &[ 2., 1.],
//! &[ 3., 2.],
//! ]);
//! let y = vec![1., 1., 1., 2., 2., 2.];
//!
//! let nb = GaussianNB::fit(&x, &y, Default::default()).unwrap();
//! let y_hat = nb.predict(&x).unwrap();
//! ```
use crate::error::Failed;
use crate::linalg::row_iter;
use crate::linalg::BaseVector;
src/naive_bayes/mod.rs
+41 -9
View File
@@ -1,3 +1,40 @@
//! # Naive Bayes
//!
//! Naive Bayes (NB) is a simple but powerful machine learning algorithm.
//! Naive Bayes classifier is based on Bayes’ Theorem with an ssumption of conditional independence
//! between every pair of features given the value of the class variable.
//!
//! Bayes’ theorem can be written as
//!
//! \\[ P(y | X) = \frac{P(y)P(X| y)}{P(X)} \\]
//!
//! where
//!
//! * \\(X = (x_1,...x_n)\\) represents the predictors.
//! * \\(P(y | X)\\) is the probability of class _y_ given the data X
//! * \\(P(X| y)\\) is the probability of data X given the class _y_.
//! * \\(P(y)\\) is the probability of class y. This is called the prior probability of y.
//! * \\(P(y | X)\\) is the probability of the data (regardless of the class value).
//!
//! The naive conditional independence assumption let us rewrite this equation as
//!
//! \\[ P(y | x_1,...x_n) = \frac{P(y)\prod_{i=1}^nP(x_i|y)}{P(x_1,...x_n)} \\]
//!
//!
//! The denominator can be removed since \\(P(x_1,...x_n)\\) is constrant for all the entries in the dataset.
//!
//! \\[ P(y | x_1,...x_n) \propto P(y)\prod_{i=1}^nP(x_i|y) \\]
//!
//! To find class y from predictors X we use this equation
//!
//! \\[ y = \underset{y}{argmax} P(y)\prod_{i=1}^nP(x_i|y) \\]
//!
//! ## References:
//!
//! * ["Machine Learning: A Probabilistic Perspective", Kevin P. Murphy, 2012, Chapter 3 ](https://mitpress.mit.edu/books/machine-learning-1)
//!
//! <script src="https://polyfill.io/v3/polyfill.min.js?features=es6"></script>
//! <script id="MathJax-script" async src="https://cdn.jsdelivr.net/npm/mathjax@3/es5/tex-mml-chtml.js"></script>
use crate::error::Failed;
use crate::linalg::BaseVector;
use crate::linalg::Matrix;
@@ -64,12 +101,7 @@ impl<T: RealNumber, M: Matrix<T>, D: NBDistribution<T, M>> BaseNaiveBayes<T, M,
Ok(y_hat)
}
}
mod bernoulli;
mod categorical;
mod gaussian;
mod multinomial;
pub use bernoulli::{BernoulliNB, BernoulliNBParameters};
pub use categorical::{CategoricalNB, CategoricalNBParameters};
pub use gaussian::{GaussianNB, GaussianNBParameters};
pub use multinomial::{MultinomialNB, MultinomialNBParameters};
pub mod bernoulli;
pub mod categorical;
pub mod gaussian;
pub mod multinomial;
src/naive_bayes/multinomial.rs
+35
View File
@@ -1,3 +1,38 @@
//! # Multinomial Naive Bayes
//!
//! Multinomial Naive Bayes classifier is a variant of [Naive Bayes](../index.html) for the multinomially distributed data.
//! It is often used for discrete data with predictors representing the number of times an event was observed in a particular instance,
//! for example frequency of the words present in the document.
//!
//! Example:
//!
//! ```
//! use smartcore::linalg::naive::dense_matrix::*;
//! use smartcore::naive_bayes::multinomial::MultinomialNB;
//!
//! // Training data points are:
//! // Chinese Beijing Chinese (class: China)
//! // Chinese Chinese Shanghai (class: China)
//! // Chinese Macao (class: China)
//! // Tokyo Japan Chinese (class: Japan)
//! let x = DenseMatrix::<f64>::from_2d_array(&[
//! &[1., 2., 0., 0., 0., 0.],
//! &[0., 2., 0., 0., 1., 0.],
//! &[0., 1., 0., 1., 0., 0.],
//! &[0., 1., 1., 0., 0., 1.],
//! ]);
//! let y = vec![0., 0., 0., 1.];
//! let nb = MultinomialNB::fit(&x, &y, Default::default()).unwrap();
//!
//! // Testing data point is:
//! // Chinese Chinese Chinese Tokyo Japan
//! let x_test = DenseMatrix::<f64>::from_2d_array(&[&[0., 3., 1., 0., 0., 1.]]);
//! let y_hat = nb.predict(&x_test).unwrap();
//! ```
//!
//! ## References:
//!
//! * ["Introduction to Information Retrieval", Manning C. D., Raghavan P., Schutze H., 2009, Chapter 13 ](https://nlp.stanford.edu/IR-book/information-retrieval-book.html)
use crate::error::Failed;
use crate::linalg::row_iter;
use crate::linalg::BaseVector;
src/svm/svc.rs
-1
View File
@@ -28,7 +28,6 @@
//!
//! ```
//! use smartcore::linalg::naive::dense_matrix::*;
//! use smartcore::linear::linear_regression::*;
//! use smartcore::svm::Kernels;
//! use smartcore::svm::svc::{SVC, SVCParameters};
//!