APPARATUS AND METHOD FOR ANALYZING DATA

摘要

In data analysis for determining a correlation model among a condition of a living body and the expression doses of plural genes and/or the amount of an intracellular substance, an explaining variable contained in the data is selected in a data mass wherein the condition of the living body or a change in the living body probabilistically occurring with the passage of time is regarded as a target variable while the expression doses of plural genes and/or the amount of an intracellular substance are regarded as explaining variables. Then the cross verification of a correlation model containing the thus selected explaining variable and target variable is calculated and the results are evaluated. The selection of the explaining variable, the calculation of cross verification and the evaluation of the results are repeated until the cross verification is not improved any more to thereby determine a partial least squares method model. Thus, an effective data processing method for multivariable gene expression data is provided.