Some Pooling Designs About Complex

In this paper, we generalize the result which is derived in the condition that the test outcomeis ideal in pooling design to the error-tolerant version. And we generalize the result which is derived in the condition that the positive element is single to the version that the positive element is k₁- complex (k₁≤k). Firstly, we propose the definition ofα-almost (k; 2e + 1)-separable matrix on the base ofα-almost k-disjunct matrix. We use the matrix which results by the complement of theα-almost (k; 2e + 1) - separable matrix intersecting random rows to construct an error-tolerant pooling design for k- complexes. And we calculate the expected number of identifying all positive k - complexes. Next, we propose the definition ofα-almost k^e-disjunct matrix. We use the matrix which results by the complement of theα-almost k^e-disjunct matrix intersecting random rows to construct an error-tolerant pooling design for k₁ - complexes. And we calculate the expected number of identifying all positive k₁- complexes. Finally, we use the matrix of (s, l)-superimposed code to give a pooling design which fits all inhibitor models for k₁ - complexes. We use the matrix of (s, l)^e-cover free family to give an error-tolerant pooling design in the inhibitor model for k₁ - complexes. we propose the definition of (k, d + (m out of r))-superimposed code on the base of (d + (m out of r))-disjunct matrix. We use the matrix of (k,d + (m out of r))-superimposed code to give a pooling design for k₁- complexes.