Unifying back-propagation and forward-forward algorithms through model predictive control
This work addresses a foundational issue in machine learning by providing a unified training framework, though it appears incremental as it builds on existing algorithms.
The paper tackles the problem of unifying back-propagation and forward-forward algorithms for training deep neural networks by introducing a Model Predictive Control framework, resulting in a range of intermediate algorithms with a performance-efficiency trade-off and a principled method for choosing optimization horizons, demonstrated through numerical results on various models and tasks.
We introduce a Model Predictive Control (MPC) framework for training deep neural networks, systematically unifying the Back-Propagation (BP) and Forward-Forward (FF) algorithms. At the same time, it gives rise to a range of intermediate training algorithms with varying look-forward horizons, leading to a performance-efficiency trade-off. We perform a precise analysis of this trade-off on a deep linear network, where the qualitative conclusions carry over to general networks. Based on our analysis, we propose a principled method to choose the optimization horizon based on given objectives and model specifications. Numerical results on various models and tasks demonstrate the versatility of our method.