The increasing demand for cleaner combustion and reduced greenhouse gas emissions motivates research on the combustion of hydrocarbon fuels and their surrogates. Accurate detailed chemical kinetic models are an important prerequisite for high fidelity reacting flow simulations capable of improving combustor design and operation. The development of such models for many new fuel components and/or surrogate molecules is greatly facilitated by the application of reaction classes and rate rules. Accurate and versatile rate rules are desirable to improve the predictive accuracy of kinetic models. A major contribution in the literature is the recent work by Bugler et al. (2015), which has significantly improved rate rules and thermochemical parameters used in kinetic modeling of alkanes. In the present study, it is demonstrated that rate rules can be used and consistently optimized for a set of normal alkanes including n-heptane, n-octane, n-nonane, n-decane, and n-undecane, thereby improving the predictive accuracy for all the considered fuels. A Bayesian framework is applied in the calibration of the rate rules. The optimized rate rules are subsequently applied to generate a mechanism for n-dodecane, which was not part of the training set for the optimized rate rules. The developed mechanism shows accurate predictions compared with published well-validated mechanisms for a wide range of conditions.