Two Losses Are Better Than One: Faster Optimization Using a Cheaper Proxy
This provides a principled method for leveraging synthetic data or simulators in machine learning, though it is incremental as it builds on existing optimization techniques.
The paper tackles the problem of optimizing objectives with hard-to-compute gradients by using a cheaper proxy function, achieving convergence rates matching stochastic gradient descent on smooth objectives, which improves sample efficiency.
We present an algorithm for minimizing an objective with hard-to-compute gradients by using a related, easier-to-access function as a proxy. Our algorithm is based on approximate proximal point iterations on the proxy combined with relatively few stochastic gradients from the objective. When the difference between the objective and the proxy is $δ$-smooth, our algorithm guarantees convergence at a rate matching stochastic gradient descent on a $δ$-smooth objective, which can lead to substantially better sample efficiency. Our algorithm has many potential applications in machine learning, and provides a principled means of leveraging synthetic data, physics simulators, mixed public and private data, and more.