The present work deals with the application of the golden section search method (GSSM) for predicting the internal rate of heat generation to reconstruct a given temperature distribution within a rectangular fin involving all modes of heat transfer. The thermal conductivity has been assumed to be temperature-dependent. The forward problem is numerically solved using an implicit fourth-order Runge-Kutta method, whereas, an inverse problem has been solved using GSSM. In conjunction with GSSM, for the inverse analysis, the effect of inverse crime has been addressed using a different solver operating on fifth-order accurate Runge-Kutta method than that used for synthesizing the input data. A case study of Hastelloy generally used in gas turbine applications is also presented and the effect of measurement error in the temperature distribution has been reported. For pure temperature data, an exact estimation of the internal heat generation rate is done, whereas, even with noisy data, a satisfactory estimation of the heat generation rate is also achieved which is verified from the reconstructed temperature distributions.