Source code for prosemble.models.sng

"""
Supervised Neural Gas (SNG).

Combines GLVQ's classification loss with Neural Gas neighborhood
cooperation using plain squared Euclidean distance. Unlike SRNG,
SNG does not adapt per-feature relevance weights — it operates
purely in the input space with isotropic distance.

References
----------
.. [1] Hammer, B., Strickert, M., & Villmann, T. (2005). Supervised
       Neural Gas and Extensions. In Proceedings of the Workshop on
       New Challenges in Neural Computation (NC2).
"""

import jax
import jax.numpy as jnp
import numpy as np

from prosemble.models.prototype_base import SupervisedPrototypeModel


[docs] class SNG(SupervisedPrototypeModel): """Supervised Neural Gas. Combines two key ideas: - GLVQ loss: :math:`(d^+ - d^-) / (d^+ + d^-)` for margin-based classification - Neural Gas cooperation: all same-class prototypes participate in the loss, weighted by rank via :math:`\\exp(-\\text{rank} / \\gamma)` Uses plain squared Euclidean distance without any metric adaptation. The neighborhood range :math:`\\gamma` decays during training from :math:`\\gamma_{\\text{init}}` to :math:`\\gamma_{\\text{final}}`. When :math:`\\gamma \\to 0`, SNG recovers standard GLVQ. Parameters ---------- beta : float Transfer function steepness parameter for sigmoid shaping. gamma_init : float, optional Initial neighborhood range for NG cooperation. Default: max prototypes per class / 2. gamma_final : float Final neighborhood range. Default: 0.01. gamma_decay : float, optional Per-step multiplicative decay factor for gamma. Default: computed from max_iter so gamma reaches gamma_final. lr_ratio : float Ratio of wrong-class to correct-class learning rate (ε⁻/ε⁺). Default: 0.5. 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. batch_size : int, optional Mini-batch size. If None (default), use full-batch training. lr_scheduler : str or optax.Schedule, optional Learning rate schedule. lr_scheduler_kwargs : dict, optional Keyword arguments for the learning rate scheduler. prototypes_initializer : str or callable, optional How to initialize prototypes. patience : int, optional Number of consecutive epochs with no improvement before stopping. restore_best : bool If True, restore parameters that achieved lowest loss. Default: False. class_weight : dict or 'balanced', optional Weights for each class. Default: None (uniform). gradient_accumulation_steps : int, optional Accumulate gradients over this many steps before applying. ema_decay : float, optional Exponential moving average decay for parameters. freeze_params : list of str, optional Parameter group names to freeze. lookahead : dict, optional Enable lookahead optimizer wrapper. mixed_precision : str or None, optional Compute dtype for mixed precision training. """ def __init__(self, beta=10.0, gamma_init=None, gamma_final=0.01, gamma_decay=None, lr_ratio=0.5, 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): 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, ) self.beta = beta self.gamma_init = gamma_init self.gamma_final = gamma_final self.gamma_decay = gamma_decay self.lr_ratio = lr_ratio self.gamma_ = None if self.freeze_params is None: self.freeze_params = ['gamma'] elif 'gamma' not in self.freeze_params: self.freeze_params = list(self.freeze_params) + ['gamma'] def _get_resume_params(self, params): gamma = self.gamma_ if self.gamma_ is not None else ( self._gamma_init_actual if hasattr(self, '_gamma_init_actual') else 1.0 ) params['gamma'] = jnp.array(gamma, dtype=jnp.float32) return params def _init_state(self, X, y, key): key1, key2 = jax.random.split(key) prototypes, proto_labels = self._init_prototypes( X, y, self.n_prototypes_per_class, key1 ) if isinstance(self.n_prototypes_per_class, int): max_per_class = self.n_prototypes_per_class elif isinstance(self.n_prototypes_per_class, dict): max_per_class = max(self.n_prototypes_per_class.values()) else: max_per_class = max(self.n_prototypes_per_class) gamma_init = self.gamma_init if self.gamma_init is not None else max_per_class / 2.0 gamma_init = max(gamma_init, self.gamma_final + 1e-6) self._gamma_init_actual = gamma_init if self.gamma_decay is not None: self._gamma_decay = self.gamma_decay else: if self.batch_size is not None: steps_per_epoch = (X.shape[0] + self.batch_size - 1) // self.batch_size else: steps_per_epoch = 1 total_steps = self.max_iter * steps_per_epoch self._gamma_decay = (self.gamma_final / gamma_init) ** (1.0 / total_steps) params = { 'prototypes': prototypes, 'gamma': jnp.array(gamma_init, dtype=jnp.float32), } opt_state = self._optimizer.init(params) from prosemble.models.prototype_base import SupervisedState state = SupervisedState( prototypes=prototypes, opt_state=opt_state, loss=jnp.array(float('inf')), iteration=0, converged=False, ) return state, params, proto_labels def _compute_loss(self, params, X, y, proto_labels): prototypes = params['prototypes'] gamma = params['gamma'] # Squared Euclidean distance diff = X[:, None, :] - prototypes[None, :, :] # (n, p, d) distances = jnp.sum(diff ** 2, axis=2) # (n, p) # Compute ranks within same-class prototypes same_class = (y[:, None] == proto_labels[None, :]) # (n, p) INF = jnp.finfo(distances.dtype).max d_same = jnp.where(same_class, distances, INF) # (n, p) order = jnp.argsort(d_same, axis=1) ranks = jnp.argsort(order, axis=1).astype(jnp.float32) # (n, p) # Neighborhood function h = exp(-rank / gamma) h = jnp.exp(-ranks / (gamma + 1e-10)) # (n, p) h = jnp.where(same_class, h, 0.0) # Normalize C = jnp.sum(h, axis=1, keepdims=True) # (n, 1) h_normalized = h / (C + 1e-10) # (n, p) # Closest different-class prototype distance d_diff = jnp.where(~same_class, distances, INF) dm = jnp.min(d_diff, axis=1) # (n,) # Separate learning rates dm = jax.lax.stop_gradient(dm) + self.lr_ratio * ( dm - jax.lax.stop_gradient(dm)) # GLVQ mu mu = (distances - dm[:, None]) / (distances + dm[:, None] + 1e-10) # (n, p) # Transfer function from prosemble.core.activations import sigmoid_beta transfer = self.transfer_fn or sigmoid_beta cost = transfer(mu + self.margin, self.beta) # (n, p) # Rank-weighted sum weighted_cost = jnp.sum(h_normalized * cost, axis=1) # (n,) return jnp.mean(weighted_cost) def _post_update(self, params): new_gamma = params['gamma'] * self._gamma_decay new_gamma = jnp.maximum(new_gamma, self.gamma_final) return {**params, 'gamma': new_gamma} def _extract_results(self, params, proto_labels, loss_history, n_iter, **kwargs): super()._extract_results(params, proto_labels, loss_history, n_iter, **kwargs) self.gamma_ = float(params['gamma']) def _get_fitted_arrays(self): arrays = super()._get_fitted_arrays() if self.gamma_ is not None: arrays['gamma_'] = np.asarray(self.gamma_) return arrays def _set_fitted_arrays(self, arrays): super()._set_fitted_arrays(arrays) if 'gamma_' in arrays: self.gamma_ = float(arrays['gamma_']) def _get_hyperparams(self): hp = super()._get_hyperparams() hp['beta'] = self.beta hp['gamma_init'] = self.gamma_init hp['gamma_final'] = self.gamma_final hp['gamma_decay'] = self.gamma_decay hp['lr_ratio'] = self.lr_ratio return hp