Artificial Immune Recognition System (AIRS): An Immune-Inspired Supervised Learning Algorithm
read more
Citations
Machine learning
Review: The use of computational intelligence in intrusion detection systems: A review
Theoretical advances in artificial immune systems
Revisiting the Foundations of Artificial Immune Systems for Data Mining
A review of clonal selection algorithm and its applications
References
Neural Networks: A Comprehensive Foundation
Neural networks for pattern recognition
Machine learning
Neural Networks for Pattern Recognition
Related Papers (5)
Frequently Asked Questions (13)
Q2. What are the future works mentioned in the paper "Artificial immune recognition system (airs): an immune- inspired supervised learning algorithm" ?
Future work lies in the application of AIRS to an immunological problem: predicting the binding or non-binding of T-cell receptors. This work will also be extended to a seven class T-cell binding prediction, as with all things in immunology, the two class binding problem is only the first stage in a very complex classification process.
Q3. What is the primary mechanism for providing evolutionary pressure to the population of ARBs in the development?
The primary mechanism for providing evolutionary pressure to the population of ARBs in the development of memory cells is the competition for system wide resources.
Q4. What is the final stage of the training algorithm?
The final stage of the training algorithm is the potential introduction of the candidate memory cell into the set of established memory cells.
Q5. What is the role of the B-Cell in the process of cloning?
During the process of clonal expansion, the B-Cell undergoes rapid proliferation (cloning) in proportion to how well it matches the antigen.
Q6. What was the threshold for training the ARB pool on an antigen?
In order to stop training on an antigen the average normalized stimulation level had to exceed the stimulation threshold for each class group of ARBs.
Q7. What is the point of the interaction of the ARB pool with the antigenic material?
The point of the interaction of the ARB pool with the antigenic material is really only in evolving a good potential memory cell, and this potential memory cell must be of the same class as the training antigen.
Q8. How many runs were used to compare the results of the Sonar data set?
the Sonar data set utilized the thirteen-way cross validation suggested in the literature [4] and was averaged over ten runs to allow for more direct comparisons with other experiments reported in the literature.
Q9. What is the definition of a function used to measure the response of an ARB to an?
If all of a given ARB’s resources are removed, then that ARB is removed from the cell population.− seed cell : an antibody, drawn from the training set, used to initialize Memory Cell and ARB populations at the beginning of training.− stimulation function: a function used to measure the response of an ARB to an antigen or to another ARB.
Q10. What did the authors find interesting about the results of the experiment?
During the experimentation, it was noted by the authors that varying system parameters such as number of seed cells varied performance on certain data sets, however, varying system resources (i.e., the numbers of resources an ARB could compete for) seemed to have little affect.
Q11. What is the affinity between mccandidate and mcmatch?
If this test is passed, then if the affinity between mccandidate and mcmatch is less than the product of the affinity threshold and the affinity threshold scalar, then mccandidate replaces mcmatch in the set of memory cells.
Q12. What is the stopping criterion for a mutated offspring?
The only exception to this repetition is that on every pass through this portion of the algorithm, except the first pass already discussed, if the stopping criterion is met after the stimulation and resource allocation phase, then the production of mutated offspring is not performed.
Q13. What is the effect of changing the number of seed cells on classification accuracy?
In order to establish this, investigations were undertaken to determine what affect altering the number of seed cells might have on classification accuracy (figures 8 and 9) and how altering the mutation rate also might affect the classification accuracy when compared over both systems (figures 10 and 11).