485 lines
17 KiB
Rust
485 lines
17 KiB
Rust
//! # 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::api::{Predictor, SupervisedEstimator};
|
|
use crate::error::Failed;
|
|
use crate::linalg::row_iter;
|
|
use crate::linalg::BaseVector;
|
|
use crate::linalg::Matrix;
|
|
use crate::math::num::RealNumber;
|
|
use crate::math::vector::RealNumberVector;
|
|
use crate::naive_bayes::{BaseNaiveBayes, NBDistribution};
|
|
|
|
#[cfg(feature = "serde")]
|
|
use serde::{Deserialize, Serialize};
|
|
|
|
/// Naive Bayes classifier for Bearnoulli features
|
|
#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
|
|
#[derive(Debug)]
|
|
struct BernoulliNBDistribution<T: RealNumber> {
|
|
/// class labels known to the classifier
|
|
class_labels: Vec<T>,
|
|
/// number of training samples observed in each class
|
|
class_count: Vec<usize>,
|
|
/// probability of each class
|
|
class_priors: Vec<T>,
|
|
/// Number of samples encountered for each (class, feature)
|
|
feature_count: Vec<Vec<usize>>,
|
|
/// probability of features per class
|
|
feature_log_prob: Vec<Vec<T>>,
|
|
/// Number of features of each sample
|
|
n_features: usize,
|
|
}
|
|
|
|
impl<T: RealNumber> PartialEq for BernoulliNBDistribution<T> {
|
|
fn eq(&self, other: &Self) -> bool {
|
|
if self.class_labels == other.class_labels
|
|
&& self.class_count == other.class_count
|
|
&& self.class_priors == other.class_priors
|
|
&& self.feature_count == other.feature_count
|
|
&& self.n_features == other.n_features
|
|
{
|
|
for (a, b) in self
|
|
.feature_log_prob
|
|
.iter()
|
|
.zip(other.feature_log_prob.iter())
|
|
{
|
|
if !a.approximate_eq(b, T::epsilon()) {
|
|
return false;
|
|
}
|
|
}
|
|
true
|
|
} else {
|
|
false
|
|
}
|
|
}
|
|
}
|
|
|
|
impl<T: RealNumber, M: Matrix<T>> NBDistribution<T, M> for BernoulliNBDistribution<T> {
|
|
fn prior(&self, class_index: usize) -> T {
|
|
self.class_priors[class_index]
|
|
}
|
|
|
|
fn log_likelihood(&self, class_index: usize, j: &M::RowVector) -> T {
|
|
let mut likelihood = T::zero();
|
|
for feature in 0..j.len() {
|
|
let value = j.get(feature);
|
|
if value == T::one() {
|
|
likelihood += self.feature_log_prob[class_index][feature];
|
|
} else {
|
|
likelihood += (T::one() - self.feature_log_prob[class_index][feature].exp()).ln();
|
|
}
|
|
}
|
|
likelihood
|
|
}
|
|
|
|
fn classes(&self) -> &Vec<T> {
|
|
&self.class_labels
|
|
}
|
|
}
|
|
|
|
/// `BernoulliNB` parameters. Use `Default::default()` for default values.
|
|
#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
|
|
#[derive(Debug, Clone)]
|
|
pub struct BernoulliNBParameters<T: RealNumber> {
|
|
/// Additive (Laplace/Lidstone) smoothing parameter (0 for no smoothing).
|
|
pub alpha: T,
|
|
/// Prior probabilities of the classes. If specified the priors are not adjusted according to the data
|
|
pub priors: Option<Vec<T>>,
|
|
/// Threshold for binarizing (mapping to booleans) of sample features. If None, input is presumed to already consist of binary vectors.
|
|
pub binarize: Option<T>,
|
|
}
|
|
|
|
impl<T: RealNumber> BernoulliNBParameters<T> {
|
|
/// Additive (Laplace/Lidstone) smoothing parameter (0 for no smoothing).
|
|
pub fn with_alpha(mut self, alpha: T) -> Self {
|
|
self.alpha = alpha;
|
|
self
|
|
}
|
|
/// Prior probabilities of the classes. If specified the priors are not adjusted according to the data
|
|
pub fn with_priors(mut self, priors: Vec<T>) -> Self {
|
|
self.priors = Some(priors);
|
|
self
|
|
}
|
|
/// Threshold for binarizing (mapping to booleans) of sample features. If None, input is presumed to already consist of binary vectors.
|
|
pub fn with_binarize(mut self, binarize: T) -> Self {
|
|
self.binarize = Some(binarize);
|
|
self
|
|
}
|
|
}
|
|
|
|
impl<T: RealNumber> Default for BernoulliNBParameters<T> {
|
|
fn default() -> Self {
|
|
Self {
|
|
alpha: T::one(),
|
|
priors: None,
|
|
binarize: Some(T::zero()),
|
|
}
|
|
}
|
|
}
|
|
|
|
impl<T: RealNumber> BernoulliNBDistribution<T> {
|
|
/// Fits the distribution to a NxM matrix where N is number of samples and M is number of features.
|
|
/// * `x` - training data.
|
|
/// * `y` - vector with target values (classes) of length N.
|
|
/// * `priors` - Optional vector with prior probabilities of the classes. If not defined,
|
|
/// priors are adjusted according to the data.
|
|
/// * `alpha` - Additive (Laplace/Lidstone) smoothing parameter.
|
|
/// * `binarize` - Threshold for binarizing.
|
|
pub fn fit<M: Matrix<T>>(
|
|
x: &M,
|
|
y: &M::RowVector,
|
|
alpha: T,
|
|
priors: Option<Vec<T>>,
|
|
) -> Result<Self, Failed> {
|
|
let (n_samples, n_features) = x.shape();
|
|
let y_samples = y.len();
|
|
if y_samples != n_samples {
|
|
return Err(Failed::fit(&format!(
|
|
"Size of x should equal size of y; |x|=[{}], |y|=[{}]",
|
|
n_samples, y_samples
|
|
)));
|
|
}
|
|
|
|
if n_samples == 0 {
|
|
return Err(Failed::fit(&format!(
|
|
"Size of x and y should greater than 0; |x|=[{}]",
|
|
n_samples
|
|
)));
|
|
}
|
|
if alpha < T::zero() {
|
|
return Err(Failed::fit(&format!(
|
|
"Alpha should be greater than 0; |alpha|=[{}]",
|
|
alpha
|
|
)));
|
|
}
|
|
|
|
let y = y.to_vec();
|
|
|
|
let (class_labels, indices) = <Vec<T> as RealNumberVector<T>>::unique_with_indices(&y);
|
|
let mut class_count = vec![0_usize; class_labels.len()];
|
|
|
|
for class_index in indices.iter() {
|
|
class_count[*class_index] += 1;
|
|
}
|
|
|
|
let class_priors = if let Some(class_priors) = priors {
|
|
if class_priors.len() != class_labels.len() {
|
|
return Err(Failed::fit(
|
|
"Size of priors provided does not match the number of classes of the data.",
|
|
));
|
|
}
|
|
class_priors
|
|
} else {
|
|
class_count
|
|
.iter()
|
|
.map(|&c| T::from(c).unwrap() / T::from(n_samples).unwrap())
|
|
.collect()
|
|
};
|
|
|
|
let mut feature_in_class_counter = vec![vec![0_usize; n_features]; class_labels.len()];
|
|
|
|
for (row, class_index) in row_iter(x).zip(indices) {
|
|
for (idx, row_i) in row.iter().enumerate().take(n_features) {
|
|
feature_in_class_counter[class_index][idx] +=
|
|
row_i.to_usize().ok_or_else(|| {
|
|
Failed::fit(&format!(
|
|
"Elements of the matrix should be 1.0 or 0.0 |found|=[{}]",
|
|
row_i
|
|
))
|
|
})?;
|
|
}
|
|
}
|
|
|
|
let feature_log_prob = feature_in_class_counter
|
|
.iter()
|
|
.enumerate()
|
|
.map(|(class_index, feature_count)| {
|
|
feature_count
|
|
.iter()
|
|
.map(|&count| {
|
|
((T::from(count).unwrap() + alpha)
|
|
/ (T::from(class_count[class_index]).unwrap() + alpha * T::two()))
|
|
.ln()
|
|
})
|
|
.collect()
|
|
})
|
|
.collect();
|
|
|
|
Ok(Self {
|
|
class_labels,
|
|
class_priors,
|
|
class_count,
|
|
feature_count: feature_in_class_counter,
|
|
feature_log_prob,
|
|
n_features,
|
|
})
|
|
}
|
|
}
|
|
|
|
/// BernoulliNB implements the naive Bayes algorithm for data that follows the Bernoulli
|
|
/// distribution.
|
|
#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
|
|
#[derive(Debug, PartialEq)]
|
|
pub struct BernoulliNB<T: RealNumber, M: Matrix<T>> {
|
|
inner: BaseNaiveBayes<T, M, BernoulliNBDistribution<T>>,
|
|
binarize: Option<T>,
|
|
}
|
|
|
|
impl<T: RealNumber, M: Matrix<T>> SupervisedEstimator<M, M::RowVector, BernoulliNBParameters<T>>
|
|
for BernoulliNB<T, M>
|
|
{
|
|
fn fit(x: &M, y: &M::RowVector, parameters: BernoulliNBParameters<T>) -> Result<Self, Failed> {
|
|
BernoulliNB::fit(x, y, parameters)
|
|
}
|
|
}
|
|
|
|
impl<T: RealNumber, M: Matrix<T>> Predictor<M, M::RowVector> for BernoulliNB<T, M> {
|
|
fn predict(&self, x: &M) -> Result<M::RowVector, Failed> {
|
|
self.predict(x)
|
|
}
|
|
}
|
|
|
|
impl<T: RealNumber, M: Matrix<T>> BernoulliNB<T, M> {
|
|
/// Fits BernoulliNB with given data
|
|
/// * `x` - training data of size NxM where N is the number of samples and M is the number of
|
|
/// features.
|
|
/// * `y` - vector with target values (classes) of length N.
|
|
/// * `parameters` - additional parameters like class priors, alpha for smoothing and
|
|
/// binarizing threshold.
|
|
pub fn fit(
|
|
x: &M,
|
|
y: &M::RowVector,
|
|
parameters: BernoulliNBParameters<T>,
|
|
) -> Result<Self, Failed> {
|
|
let distribution = if let Some(threshold) = parameters.binarize {
|
|
BernoulliNBDistribution::fit(
|
|
&(x.binarize(threshold)),
|
|
y,
|
|
parameters.alpha,
|
|
parameters.priors,
|
|
)?
|
|
} else {
|
|
BernoulliNBDistribution::fit(x, y, parameters.alpha, parameters.priors)?
|
|
};
|
|
|
|
let inner = BaseNaiveBayes::fit(distribution)?;
|
|
Ok(Self {
|
|
inner,
|
|
binarize: parameters.binarize,
|
|
})
|
|
}
|
|
|
|
/// Estimates the class labels for the provided data.
|
|
/// * `x` - data of shape NxM where N is number of data points to estimate and M is number of features.
|
|
/// Returns a vector of size N with class estimates.
|
|
pub fn predict(&self, x: &M) -> Result<M::RowVector, Failed> {
|
|
if let Some(threshold) = self.binarize {
|
|
self.inner.predict(&(x.binarize(threshold)))
|
|
} else {
|
|
self.inner.predict(x)
|
|
}
|
|
}
|
|
|
|
/// Class labels known to the classifier.
|
|
/// Returns a vector of size n_classes.
|
|
pub fn classes(&self) -> &Vec<T> {
|
|
&self.inner.distribution.class_labels
|
|
}
|
|
|
|
/// Number of training samples observed in each class.
|
|
/// Returns a vector of size n_classes.
|
|
pub fn class_count(&self) -> &Vec<usize> {
|
|
&self.inner.distribution.class_count
|
|
}
|
|
|
|
/// Number of features of each sample
|
|
pub fn n_features(&self) -> usize {
|
|
self.inner.distribution.n_features
|
|
}
|
|
|
|
/// Number of samples encountered for each (class, feature)
|
|
/// Returns a 2d vector of shape (n_classes, n_features)
|
|
pub fn feature_count(&self) -> &Vec<Vec<usize>> {
|
|
&self.inner.distribution.feature_count
|
|
}
|
|
|
|
/// Empirical log probability of features given a class
|
|
pub fn feature_log_prob(&self) -> &Vec<Vec<T>> {
|
|
&self.inner.distribution.feature_log_prob
|
|
}
|
|
}
|
|
|
|
#[cfg(test)]
|
|
mod tests {
|
|
use super::*;
|
|
use crate::linalg::naive::dense_matrix::DenseMatrix;
|
|
|
|
#[cfg_attr(target_arch = "wasm32", wasm_bindgen_test::wasm_bindgen_test)]
|
|
#[test]
|
|
fn run_bernoulli_naive_bayes() {
|
|
// Tests that BernoulliNB when alpha=1.0 gives the same values as
|
|
// those given for the toy example in Manning, Raghavan, and
|
|
// Schuetze's "Introduction to Information Retrieval" book:
|
|
// https://nlp.stanford.edu/IR-book/html/htmledition/the-bernoulli-model-1.html
|
|
|
|
// 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 bnb = BernoulliNB::fit(&x, &y, Default::default()).unwrap();
|
|
|
|
assert_eq!(bnb.inner.distribution.class_priors, &[0.75, 0.25]);
|
|
assert_eq!(
|
|
bnb.feature_log_prob(),
|
|
&[
|
|
&[
|
|
-0.916290731874155,
|
|
-0.2231435513142097,
|
|
-1.6094379124341003,
|
|
-0.916290731874155,
|
|
-0.916290731874155,
|
|
-1.6094379124341003
|
|
],
|
|
&[
|
|
-1.0986122886681098,
|
|
-0.40546510810816444,
|
|
-0.40546510810816444,
|
|
-1.0986122886681098,
|
|
-1.0986122886681098,
|
|
-0.40546510810816444
|
|
]
|
|
]
|
|
);
|
|
|
|
// 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 = bnb.predict(&x_test).unwrap();
|
|
|
|
assert_eq!(y_hat, &[1.]);
|
|
}
|
|
|
|
#[cfg_attr(target_arch = "wasm32", wasm_bindgen_test::wasm_bindgen_test)]
|
|
#[test]
|
|
fn bernoulli_nb_scikit_parity() {
|
|
let x = DenseMatrix::<f64>::from_2d_array(&[
|
|
&[2., 4., 0., 0., 2., 1., 2., 4., 2., 0.],
|
|
&[3., 4., 0., 2., 1., 0., 1., 4., 0., 3.],
|
|
&[1., 4., 2., 4., 1., 0., 1., 2., 3., 2.],
|
|
&[0., 3., 3., 4., 1., 0., 3., 1., 1., 1.],
|
|
&[0., 2., 1., 4., 3., 4., 1., 2., 3., 1.],
|
|
&[3., 2., 4., 1., 3., 0., 2., 4., 0., 2.],
|
|
&[3., 1., 3., 0., 2., 0., 4., 4., 3., 4.],
|
|
&[2., 2., 2., 0., 1., 1., 2., 1., 0., 1.],
|
|
&[3., 3., 2., 2., 0., 2., 3., 2., 2., 3.],
|
|
&[4., 3., 4., 4., 4., 2., 2., 0., 1., 4.],
|
|
&[3., 4., 2., 2., 1., 4., 4., 4., 1., 3.],
|
|
&[3., 0., 1., 4., 4., 0., 0., 3., 2., 4.],
|
|
&[2., 0., 3., 3., 1., 2., 0., 2., 4., 1.],
|
|
&[2., 4., 0., 4., 2., 4., 1., 3., 1., 4.],
|
|
&[0., 2., 2., 3., 4., 0., 4., 4., 4., 4.],
|
|
]);
|
|
let y = vec![2., 2., 0., 0., 0., 2., 1., 1., 0., 1., 0., 0., 2., 0., 2.];
|
|
let bnb = BernoulliNB::fit(&x, &y, Default::default()).unwrap();
|
|
|
|
let y_hat = bnb.predict(&x).unwrap();
|
|
|
|
assert_eq!(bnb.classes(), &[0., 1., 2.]);
|
|
assert_eq!(bnb.class_count(), &[7, 3, 5]);
|
|
assert_eq!(bnb.n_features(), 10);
|
|
assert_eq!(
|
|
bnb.feature_count(),
|
|
&[
|
|
&[5, 6, 6, 7, 6, 4, 6, 7, 7, 7],
|
|
&[3, 3, 3, 1, 3, 2, 3, 2, 2, 3],
|
|
&[4, 4, 3, 4, 5, 2, 4, 5, 3, 4]
|
|
]
|
|
);
|
|
|
|
assert!(bnb
|
|
.inner
|
|
.distribution
|
|
.class_priors
|
|
.approximate_eq(&vec!(0.46, 0.2, 0.33), 1e-2));
|
|
assert!(bnb.feature_log_prob()[1].approximate_eq(
|
|
&vec![
|
|
-0.22314355,
|
|
-0.22314355,
|
|
-0.22314355,
|
|
-0.91629073,
|
|
-0.22314355,
|
|
-0.51082562,
|
|
-0.22314355,
|
|
-0.51082562,
|
|
-0.51082562,
|
|
-0.22314355
|
|
],
|
|
1e-1
|
|
));
|
|
assert!(y_hat.approximate_eq(
|
|
&vec!(2.0, 2.0, 0.0, 0.0, 0.0, 2.0, 1.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0),
|
|
1e-5
|
|
));
|
|
}
|
|
|
|
#[cfg_attr(target_arch = "wasm32", wasm_bindgen_test::wasm_bindgen_test)]
|
|
#[test]
|
|
#[cfg(feature = "serde")]
|
|
fn serde() {
|
|
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 bnb = BernoulliNB::fit(&x, &y, Default::default()).unwrap();
|
|
let deserialized_bnb: BernoulliNB<f64, DenseMatrix<f64>> =
|
|
serde_json::from_str(&serde_json::to_string(&bnb).unwrap()).unwrap();
|
|
|
|
assert_eq!(bnb, deserialized_bnb);
|
|
}
|
|
}
|