<p>This work investigates data-driven resilient tracking control for MASs with DoS attacks and actuator faults. The actuator faults are directly estimated with the help of the constructed interaction information-based NN. Different from the previous method, the interaction information is fully employed to update weights instead of the single agent’s information. Also, the utilization of NN approximations is avoided in the weight update law. Then, based on the I/O data of agents, a model free resilient tracking control technique is built for MASs to accomplish tracking goal, as well as resist the impacts of DoS attacks and actuator faults. The limitations of existing approaches that rely on model knowledge are overcome. Finally, the efficacy of developed technique is confirmed using simulations.</p>