Massive stars play key roles in many astrophysical processes. Deriving the atmospheric parameters of massive stars is important to understanding their physical properties, and thus the atmospheric parameters are key inputs to trace the evolution of massive stars. Here we report our work on adopting the data-driven technique called stellar label machine (SLAM) with the nonlocal thermal equilibrium TLUSTY synthetic spectra as the training data set to estimate the stellar parameters of Large Sky Area Multi-Object Fiber Spectroscopic Telescope (LAMOST) optical spectra for early-type stars. We apply two consistency tests to verify this machine-learning method and compare stellar labels given by SLAM with the labels in the literature for several objects having high-resolution spectra. We provide the stellar labels of effective temperature (T-eff), surface gravity (log g), metallicity GM/HD, and projected rotational velocity (v sin i) for 3931 and 578 early-type stars from the LAMOST low-resolution survey (LRS) and medium-resolution survey (MRS), respectively. To estimate the average statistical uncertainties of our results, we calculated the standard deviation between the predicted stellar label and the prelabeled published values from the high-resolution spectra. The uncertainties of the four parameters are sigma(T-eff) = 2185 K, sigma(log g) = 0.29 dex, and sigma(v sin i) = 11 km s(-1) for MRS, and sigma(T-eff) = 1642 K, sigma(log g) = 0.25 dex, and sigma(v sin i) = 42 km s(-1) for LRS spectra, respectively. We note that the parameters of T-eff, log g, and [M/H] can be better constrained using LRS spectra than using MRS spectra, most likely due to their broad wavelength coverage, while v sin i is constrained better by MRS spectra than by LRS spectra, probably due to the relatively accurate line profiles of MRS spectra.