Struct MLPClassifier 

#[non_exhaustive]
pub struct MLPClassifier {
    pub loss_curve: Vec<f64>,
    /* private fields */
}

Multi-layer perceptron classifier.

Trains a feedforward neural network for classification using backpropagation with configurable optimizers and activations.

Defaults match sklearn: hidden_layers=[100], Adam, lr=0.001, max_iter=200, batch_size=200, alpha=0.0001.

Fields (Non-exhaustive)

Non-exhaustive structs could have additional fields added in future. Therefore, non-exhaustive structs cannot be constructed in external crates using the traditional Struct { .. } syntax; cannot be matched against without a wildcard ..; and struct update syntax will not work.

loss_curve: Vec<f64>

Training loss curve (one entry per epoch).

Implementations

impl MLPClassifier

pub fn new() -> Self

Create a new MLP classifier with sklearn defaults.

pub fn hidden_layers(self, sizes: &[usize]) -> Self

Set hidden layer sizes. Default: &[100].

pub fn activation(self, activation: Activation) -> Self

Set activation function for hidden layers. Default: ReLU.

pub fn optimizer(self, kind: OptimizerKind) -> Self

Set optimizer algorithm. Default: Adam.

pub fn learning_rate(self, lr: f64) -> Self

Set learning rate. Default: 0.001.

pub fn max_iter(self, n: usize) -> Self

Set maximum training iterations (epochs). Default: 200.

pub fn batch_size(self, n: usize) -> Self

Set mini-batch size. Default: 200.

pub fn alpha(self, a: f64) -> Self

Set L2 regularization strength. Default: 0.0001.

pub fn tolerance(self, tol: f64) -> Self

Set convergence tolerance. Default: 1e-4.

pub fn tol(self, t: f64) -> Self

Alias for tolerance (sklearn convention).

pub fn early_stopping(self, enable: bool) -> Self

Enable early stopping with validation split. Default: false.

pub fn validation_fraction(self, frac: f64) -> Self

Set validation fraction for early stopping. Default: 0.1.

pub fn n_iter_no_change(self, n: usize) -> Self

Set patience for early stopping. Default: 10.

pub fn seed(self, s: u64) -> Self

Set random seed. Default: 42.

pub fn learning_rate_schedule(self, schedule: LearningRateSchedule) -> Self

Set learning rate schedule. Default: LearningRateSchedule::Constant.

Use LearningRateSchedule::adaptive() for reduce-on-plateau behavior.

pub fn dropout(self, p: f64) -> Self

Set dropout probability applied between hidden layers.

p is the fraction of activations to zero out (e.g. 0.5 for 50%). Applied only during training; inference is unaffected. Default: 0.0 (no dropout).

pub fn callback(self, cb: Box<dyn TrainingCallback>) -> Self

Add a training callback (invoked after each epoch).

pub fn fit(&mut self, data: &Dataset) -> Result<()>

Train the classifier on a dataset.

pub fn predict(&self, features: &[Vec<f64>]) -> Result<Vec<f64>>

Predict class labels for input samples.

features is &[Vec<f64>] where each inner vec is one sample (row-major).

pub fn predict_proba(&self, features: &[Vec<f64>]) -> Result<Vec<f64>>

Predict class probabilities (softmax output).

Returns a flat [batch * n_classes] row-major probability matrix.

pub fn n_classes(&self) -> usize

Number of classes discovered during fit.

pub fn n_features(&self) -> usize

Number of features the model was trained on.

pub fn loss_curve(&self) -> &[f64]

Training loss per epoch.

pub fn history(&self) -> Option<&TrainingHistory>

Structured training history with per-epoch metrics.

Returns None if the model has not been fitted yet.

pub fn weights(&self) -> &[(Vec<f64>, Vec<f64>)]

Saved network weights (for visualization).

pub fn layer_dims(&self) -> &[(usize, usize)]

Layer dimensions (for visualization).

pub fn activation_fn(&self) -> Activation

Hidden-layer activation function.

Trait Implementations

impl Clone for MLPClassifier

fn clone(&self) -> Self

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for MLPClassifier

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for MLPClassifier

fn default() -> Self

Returns the “default value” for a type.

impl PartialFit for MLPClassifier

fn partial_fit(&mut self, data: &Dataset) -> Result<()>

Run one epoch of mini-batch SGD on the given data.

On the first call, initializes the network architecture from the data dimensions. Subsequent calls preserve network weights and continue training.

fn is_initialized(&self) -> bool

Whether the model has been initialized (at least one partial_fit call).

impl PipelineModel for MLPClassifier

fn fit(&mut self, data: &Dataset) -> Result<()>

Train the model on a dataset.

fn predict(&self, features: &[Vec<f64>]) -> Result<Vec<f64>>

Predict on row-major feature matrix.

impl Tunable for MLPClassifier

fn set_param(&mut self, name: &str, _value: ParamValue) -> Result<()>

Apply a named hyperparameter.

fn clone_box(&self) -> Box<dyn Tunable>

Clone this model into a boxed trait object.

fn fit(&mut self, data: &Dataset) -> Result<()>

Train on a dataset.

fn predict(&self, features: &[Vec<f64>]) -> Result<Vec<f64>>

Predict on row-major features.