<p>Composite structures in their diverse sorts could enter critical industries and bring significant functions with them. Compressive loads are one of the most common types of loads that are applied to these structures, especially composite plates, in various applications, which are inclined to buckling. Therefore, in this research, various machine learning algorithms have been used to rapidly and accurately predict the buckling load of composite plates with different dimensions and boundary conditions. By employing the finite element method, 10,507 composite plate samples were analyzed using an eigenvalue analysis, and in this way, by calculating their buckling loads, the required dataset was obtained. The obtained buckling loads, the aspect ratio of the plates, and boundary conditions were given as input to machine learning models covering support vector regression, stochastic gradient descent regressor, random forest, k-nearest neighbor, and decision tree to predict the buckling load for plates with new boundary conditions that are not present in the dataset, and to discover the best model. The result was as follows that within the existing boundary conditions in the dataset, k-nearest neighbors model gives the most excellent performance with the least error, but for predicting the buckling load within the new boundary conditions, support vector regression model has performed superior.</p>