AIRS

Artificial Immune Recognition System (AIRS)

Inheritance

This class extends BaseClassifier

Module: aisp.csa
Import: from aisp.csa import AIRS

---

Overview

The Artificial Immune Recognition System (AIRS) is a classification algorithm inspired by the clonal selection process of the biological immune system. This implementation is based on the simplified AIRS2 version described in ¹. The algorithm has been adapted to support both real-valued (continuous) and binary feature datasets.

note

This implementation is inspired by AIRS2, a simplified version of the original AIRS algorithm. Introducing adaptations to handle continuous and binary datasets.

Based on Algorithm 16.5 from Brabazon et al. ¹

Related and noteworthy works: access here ².

---

Example

import numpy as np
from aisp.csa import AIRS

np.random.seed(1)
# Generating training data
a = np.random.uniform(high=0.5, size=(50, 2))
b = np.random.uniform(low=0.51, size=(50, 2))
x_train = np.vstack((a, b))
y_train = [0] * 50 + [1] * 50
# AIRS Instance
airs = AIRS(n_resources=5, rate_clonal=5, rate_hypermutation=0.65, seed=1)
airs = airs.fit(x_train, y_train, verbose=False)
x_test = [
    [0.15, 0.45],  # Expected: Class 0
    [0.85, 0.65],  # Esperado: Classe 1
]
y_pred = airs.predict(x_test)
print(y_pred)

Output:

[0 1]

---

Constructor Parameters

Name	Type	Default	Description
n_resources	float	10	Total amount of available resources.
rate_clonal	int	10	Maximum number of possible clones of a class. This quantity is multiplied by (cell_stimulus * rate_hypermutation) to define the number of clones.
rate_mc_init	float	0.2	Percentage of samples used to initialize memory cells.
rate_hypermutation	float	0.75	The rate of mutated clones derived from rate_clonal as a scalar factor.
affinity_threshold_scalar	float	0.75	Normalized affinity threshold.
k	int	3	The number of K nearest neighbors that will be used to choose a label in the prediction.
max_iters	int	100	Maximum number of iterations in the refinement process of the ARB set exposed to aᵢ.
resource_amplified	float	1.0	Resource consumption amplifier is multiplied with the incentive to subtract resources.
metric	str	"euclidean"	Distance metric used to compute affinity between cells and samples.
seed	int	None	Seed for random generation.
p	float	2	This parameter stores the value of p used in the Minkowski distance.

Attributes

Name	Type	Default	Description
cells_memory	Optional[Dict[str | int, list[BCell]]]	-	Dictionary of trained memory cells, organized by class.

---

Public Methods

fit

def fit(
    self,
    X: Union[npt.NDArray, list],
    y: Union[npt.NDArray, list],
    verbose: bool = True,
) -> AIRS:
    ...

Fit the model to the training data using the AIRS.

The function fit(...), performs the training according to X and y, using the method AIRS.

Parameters

Name	Type	Default	Description
X	Union[npt.NDArray, list]	-	Training input samples. Each row corresponds to a samples and column to feature.
y	Union[npt.NDArray, list]	-	Target vector of shape (n_samples,). Must contain the same number of samples as X.
verbose	bool	True	If True, prints training progress information.

Returns

Type	Description
Self	Returns the instance itself.

Raises

Exception	Description
TypeError	If X or y are not ndarrays or have incompatible shapes.

---

predict

def predict(self, X: Union[npt.NDArray, list]) -> npt.NDArray:
    ...

Predict class labels based on the memory cells created during training.

This method uses the trained memory cells to perform classification of the input data using the k-nearest neighbors approach.

Parameters

Name	Type	Default	Description
X	Union[npt.NDArray, list]	-	Input samples. Must have the same number of features used during training.

Returns

Type	Description
npt.NDArray	An ndarray of the form C (n_samples), containing the predicted classes for X.

Raises

Exception	Description
TypeError	If X is not a ndarray or list.
FeatureDimensionMismatch	If the number of features in X does not match the expected number.
ModelNotFittedError	If the mode has not yet been adjusted and does not have defined memory cells, it is not able to predictions

---

Extended Example

Complete usage examples are available in the Jupyter Notebooks:

• Examples

---

References

Footnotes

1. Brabazon, A., O'Neill, M., & McGarraghy, S. (2015). Natural Computing Algorithms. In Natural Computing Series. Springer Berlin Heidelberg. https://doi.org/10.1007/978-3-662-43631-8 ↩ ↩²

2. AZZOUG, Aghiles. Artificial Immune Recognition System V2. Available at: https://github.com/AghilesAzzoug/Artificial-Immune-System ↩