RBF network modeling method based on immune polyclonal optimization in DNA sequence classification

Abstract

The invention discloses an RBF (Radial Basis Function) network modeling method based on immune polyclonal optimization in DNA (Desoxvribose Nucleic Acid) sequence classification. The RBF network modeling method comprises the following steps of randomly generating an initial antibody population A={a1, a2,..., an}, calculating the affinity function f(x) of the antibodies in the antibody population, putting the antibodies in the antibody population in a descending order according to the values of f(x) to obtain A'={a'1, a'2,..., a'N}, selecting m antibodies, each of which the affinity function f(x) has a greater value, from A', and performing cloning operation on the m antibodies to obtain a new antibody population A', performing clonal variation and clonal crossover operations on the current population A'', respectively, to obtain a new population FORMULA, performing clonal selection operation FORMULA on the population FORMULA, outputting the antibody which simultaneously satisfies the conditions of the minimum support and the minimum confidence, and in the meantime, reducing the antibody into the primitive attribute value and remaining the antibody in the population, and when k is greater than or equal to Genmax, finishing the algorithm and completing molding, otherwise, determining that k is equal to k+1, taking the present population as the initial antibody population for the calculation of next generation and turning to the step 2. As a result, the purposes of improving the DNA sequence classification efficiency and improving the reserve ratio are achieved.

Description

technical field [0001] The invention relates to the field of genetic information classification, in particular to an RBF network modeling method based on immune polyclonal optimization in DNA sequence classification. Background technique [0002] At present, it is an important direction of data mining research to study the hidden rules in the sequence composed of four bases: A, T, C, and G in the complete DNA sequence. Although humans know little about this structure at this stage, they have discovered some regularities in the DNA sequence. Some scholars at home and abroad have used data mining technology to conduct related research on DNA sequences, and have carried out mining research on DNA sequences from the perspectives of support vector machine (vsm model), BP network, cluster analysis and genetic algorithm, but the final effect Not as expected. [0003] The RBF network is a widely used feed-forward neural network. The structure of the neural network is as follows: ...

AI Technical Summary

Problems solved by technology

The Gaussian function used in the radial basis network is to optimize the network weights by the gradient descent method. The main problem of this method is that it is easy to mature and the training time is long.

resulting in a large amount of calculation

Images