Machine learning (ML) has previously been applied to predict reaction barriers for a variety of different chemical reactions. This is seen as the end point for this type of study however, post-reaction barrier analysis/energy decomposition approaches can provide insight into chemical reactivity. One such approach that has previously been used to provide information on chemical reactivity, for cycloaddition reactions in particular, is distortion/interaction-activation strain analysis (DIAS). We demonstrate that ML can be coupled with cheap and rapid semi-empirical quantum mechanical methods (SQM) to predict distortion and interaction energies at a fraction of the computational cost associated with running density functional theory (DFT) calculations.  This dataset includes all the structural data in the form of Gaussian16 (Revision A.03 and C.01) output files for the four datasets used in this work and, the literature dataset reactions.