Useful disturbance: How non-linear dynamics can increase edge sensor time collection

Engineers have demonstrated a easy computational method for supporting the classification efficiency of neural networks working on sensor time collection. The proposed approach includes feeding the recorded sign as an exterior forcing into an elementary non-linear dynamical system, and offering its temporal responses to this disturbance to the neural community alongside the unique knowledge.