<p>This study proposes a capturability assessment algorithm to quantify the interception performance against targets across a diverse spectrum of speeds, including those slower and faster than the interceptor. To represent the engagement scenario, a relative velocity-based space is utilized, and the adequacy of such representation in fully describing the given engagement scenario is theoretically established. Furthermore, the range of feasible initial conditions for interception, derived from the relative velocity space, is employed as an interception performance metric. A data-driven analytical model using Gaussian Process Regression is developed to predict capturability while accounting for uncertainties in target acceleration; this model is efficiently constructed using a theoretically justified boundary-biased sampling scheme. The effectiveness of the proposed analytical techniques is validated through numerical simulations, which cover a wide range of target speeds and maneuvering profiles.</p>