The immune system as a model for pattern recognition and classification

Objective: To design a pattern recognition engine based on concepts derived from mammalian immune systems. Design: A supervised learning system (Immunos-81) was created using softwar e abstractions of T cells, B cells, antibodies, and their interactions. Art ificial T cells control the creation of a-cell populations (clones), which compete for recognition of "unknowns." The B-cell clone with the "simple hi ghest avidity" (SHA) or "relative highest avidity" (RHA) is considered to h ave successfully classified the unknown. Measurement: Two standard machine learning data sets, consisting of eight n ominal and six continuous variables, were used to test the recognition capa bilities of Immunos-81. The first set (Cleveland), consisting of 303 cases of patients with suspected coronary artery disease, was used to perform a t en-way cross-validation. After completing the validation runs, the Clevelan d data set was used-as-a training set prior to presentation of the second d ata set, consisting of 200 unknown cases. Results: For cross-validation runs, correct recognition using SHA ranged fr om a high of 96 percent to a low of 63.2 percent. The average correct class ification for all runs was 83.2 percent. Using the RHA metric, 11.2 percent were labeled '"too close to determine" and net further attempt was made to classify them. Of the remaining cases, 85.5 percent were correctly classif ied. When the second data set was presented, correct classification occurre d in 73.5 percent of cases when SHA was used and in 80.3 percent of cases w hen RHA was used. Conclusions: The immune system offers a viable paradigm for the design of p attern recognition systems. Additional research is required to fully exploi t the nuances of immune computation.