Given a sequence of observable variables , the conformal prediction methodology estimates a confidence set for given that’s legitimate for any finite pattern measurement by merely assuming that the joint distribution of the information is permutation invariant. Though engaging, computing such a set is computationally infeasible in most regression issues. Certainly, in these circumstances, the unknown variable can take an infinite variety of potential candidate values, and producing conformal units requires retraining a predictive mannequin for every candidate. On this paper, we give attention to a sparse linear mannequin with solely a subset of variables for prediction and use numerical continuation methods to approximate the answer path effectively. The essential property we exploit is that the set of chosen variables is invariant beneath a small perturbation of the enter information. Due to this fact, it’s ample to enumerate and refit the mannequin solely on the change factors of the set of energetic options and easily interpolate the remainder of the answer through a Predictor-Corrector mechanism. We present how our path-following algorithm precisely approximates conformal prediction units and illustrate its efficiency utilizing artificial and actual information examples.