"""
Generalized Learning Vector Quantization (GLVQ) and variants.
Implements GLVQ, GLVQ1, and GLVQ21 — all gradient-based supervised
prototype models using the relative distance difference loss.
References
----------
.. [1] Sato, A., & Yamada, K. (1995). Generalized Learning Vector
Quantization. NIPS.
"""
import jax
import jax.numpy as jnp
from prosemble.models.prototype_base import SupervisedPrototypeModel
from prosemble.core.losses import glvq_loss, glvq_loss_with_transfer, lvq1_loss, lvq21_loss
[docs]
class GLVQ(SupervisedPrototypeModel):
"""Generalized Learning Vector Quantization.
Loss:
.. math::
\\mu = \\frac{d^+ - d^-}{d^+ + d^-}
with optional transfer function.
Parameters
----------
beta : float
Parameter :math:`\\beta` for transfer function (e.g., sigmoid steepness).
n_prototypes_per_class : int
Number of prototypes per class.
max_iter : int
Maximum training iterations.
lr : float
Learning rate.
epsilon : float
Convergence threshold on loss change.
random_seed : int
Random seed for reproducibility.
distance_fn : callable, optional
Distance function (default: squared Euclidean).
optimizer : str or optax optimizer, optional
Optimizer name ('adam', 'sgd') or optax GradientTransformation.
Default: 'adam'.
transfer_fn : callable, optional
Transfer function for loss shaping (default: identity).
margin : float
Margin for loss computation.
callbacks : list, optional
List of Callback objects.
use_scan : bool
If True (default), use jax.lax.scan for training (faster, JIT-compiled,
but runs all max_iter iterations even after convergence).
If False, use a Python for-loop with true early stopping (no wasted
compute after convergence, but slower per iteration).
batch_size : int, optional
Mini-batch size. If None (default), use full-batch training.
When set, each epoch iterates over shuffled mini-batches of this size.
lr_scheduler : str or optax.Schedule, optional
Learning rate schedule. Supported strings: 'exponential_decay',
'cosine_decay', 'warmup_cosine_decay', 'warmup_exponential_decay',
'warmup_constant', 'polynomial', 'linear', 'piecewise_constant',
'sgdr'. Or pass a custom optax.Schedule. Default: None.
lr_scheduler_kwargs : dict, optional
Keyword arguments passed to the learning rate scheduler
(e.g. ``decay_rate``, ``transition_steps``). Default: None.
prototypes_initializer : str or callable, optional
How to initialize prototypes. Supported strings: 'stratified_random'
(default), 'class_mean', 'class_conditional_mean', 'stratified_noise',
'random_normal', 'uniform', 'zeros', 'ones', 'fill_value'.
Or pass a callable ``(X, y, n_per_class, key) -> (protos, labels)``.
patience : int, optional
Number of consecutive epochs with no improvement before stopping.
If None (default), stops after a single non-improving step (epsilon
check). Requires use_scan=False for true early stopping.
restore_best : bool
If True, restore the parameters that achieved the lowest loss
(or validation loss if validation data is provided). Default: False.
class_weight : dict or 'balanced', optional
Weights for each class. Dict maps class label to weight, e.g.
{0: 1.0, 1: 2.0, 2: 1.5}. 'balanced' auto-computes weights
inversely proportional to class frequencies. Default: None (uniform).
gradient_accumulation_steps : int, optional
Accumulate gradients over this many steps before applying an update.
Effective batch size = batch_size * gradient_accumulation_steps.
Default: None (no accumulation).
ema_decay : float, optional
Exponential moving average decay for parameters (0 < ema_decay < 1).
After training, model parameters are replaced with EMA-smoothed values.
Typical values: 0.999, 0.9999. Default: None (no EMA).
freeze_params : list of str, optional
List of parameter group names to freeze (zero gradients).
E.g. ['backbone'] to freeze the backbone and only train prototypes.
Default: None (all parameters trainable).
lookahead : dict, optional
Enable lookahead optimizer wrapper. Dict with keys:
- 'sync_period': int (default 6) — sync every k steps
- 'slow_step_size': float (default 0.5) — interpolation factor
Default: None (no lookahead).
mixed_precision : str or None, optional
Compute dtype for mixed precision training. 'float16' or 'bfloat16'.
Master weights stay in float32; forward/backward pass runs in lower
precision for ~2x speed and ~half memory on GPU. Float16 uses static
loss scaling to prevent gradient underflow. Default: None (disabled).
"""
def __init__(self, beta=10.0, n_prototypes_per_class=1, max_iter=100,
lr=0.01, epsilon=1e-6, random_seed=42, distance_fn=None,
optimizer='adam', transfer_fn=None, margin=0.0,
callbacks=None, use_scan=True, batch_size=None,
lr_scheduler=None, lr_scheduler_kwargs=None,
prototypes_initializer=None, patience=None,
restore_best=False, class_weight=None,
gradient_accumulation_steps=None, ema_decay=None,
freeze_params=None, lookahead=None, mixed_precision=None,
**kwargs):
super().__init__(
n_prototypes_per_class=n_prototypes_per_class,
max_iter=max_iter, lr=lr, epsilon=epsilon,
random_seed=random_seed, distance_fn=distance_fn,
optimizer=optimizer, transfer_fn=transfer_fn, margin=margin,
callbacks=callbacks, use_scan=use_scan, batch_size=batch_size,
lr_scheduler=lr_scheduler,
lr_scheduler_kwargs=lr_scheduler_kwargs,
prototypes_initializer=prototypes_initializer,
patience=patience, restore_best=restore_best,
class_weight=class_weight,
gradient_accumulation_steps=gradient_accumulation_steps,
ema_decay=ema_decay, freeze_params=freeze_params,
lookahead=lookahead, mixed_precision=mixed_precision,
**kwargs,
)
self.beta = beta
def _compute_loss(self, params, X, y, proto_labels):
distances = self.distance_fn(X, params['prototypes'])
return glvq_loss_with_transfer(
distances, y, proto_labels,
transfer_fn=self.transfer_fn,
margin=self.margin,
beta=self.beta,
)
def _get_hyperparams(self):
hp = super()._get_hyperparams()
hp['beta'] = self.beta
return hp
[docs]
class GLVQ1(SupervisedPrototypeModel):
"""GLVQ with LVQ1-style loss (gradient-based).
Loss: :math:`d^+` when correct, :math:`-d^-` when wrong.
Parameters
----------
n_prototypes_per_class : int
Prototypes per class.
max_iter : int
Maximum training iterations.
lr : float
Learning rate.
See Also
--------
SupervisedPrototypeModel : Full list of base parameters (optimizer,
distance_fn, lr_scheduler, callbacks, patience, etc.).
"""
def _compute_loss(self, params, X, y, proto_labels):
distances = self.distance_fn(X, params['prototypes'])
return lvq1_loss(distances, y, proto_labels)
[docs]
class GLVQ21(SupervisedPrototypeModel):
"""GLVQ with LVQ2.1-style loss (gradient-based, unnormalized).
Loss: :math:`d^+ - d^-` (no normalization by :math:`d^+ + d^-`).
Parameters
----------
n_prototypes_per_class : int
Prototypes per class.
max_iter : int
Maximum training iterations.
lr : float
Learning rate.
See Also
--------
SupervisedPrototypeModel : Full list of base parameters (optimizer,
distance_fn, lr_scheduler, callbacks, patience, etc.).
"""
def _compute_loss(self, params, X, y, proto_labels):
distances = self.distance_fn(X, params['prototypes'])
return lvq21_loss(distances, y, proto_labels)