Vector monitoring remains a challenging task for large-scale continuous surveillance. Light Detection and Ranging (Lidar) systems, which use laser pulses to detect and characterize objects at a distance (Lidar systems), offer a solution by capturing the wingbeat modulation frequencies of flying insects, especially mosquitoes. However, the wingbeat frequencies of different mosquito species are often very similar. The use of additional parameters is necessary to improve the classification. In this study, we used a Scheimpflug lidar system to record the kHz-modulated backscattered light of mosquitoes and applied the Random Forest method to select the most discriminating parameters. The simplified model, based only on two parameters, the fundamental frequency (Freq) and the first harmonic (Harm1), achieved an accuracy of 44.14%, as shown in the confusion matrix. Although this accuracy may seem modest, it is remarkable considering the inherent challenges of the problem, such as the reduced parameter set and the complexity of the data. These parameters enabled better separation of mosquito specimens collected by the lidar, demonstrating that reducing the number of parameters can not only maintain but also improve the model's classification accuracy while reducing its complexity. Keywords: Scheimpflug Lidar, Wingbeat Frequency, Harmonic, Mosquito Classification