서명환 교수의 논문 “Fast and Robust Online Inference with Stochastic Gradient Descent via Random Scaling”, coauthored with Sokbae Lee, Yuan Liao, and Youngki Shin, 이 AAAI-2022, the Thirty-sixth AAAI Conference on Artificial Intelligence, 학회에 채택되었다.

The Association For The Advancement Of Artificial Intelligence (AAAI)는 인공지능 분야에서 가장 권위 있는 학술대회 중 하나로, 학회는 2022년 2월 22일부터 3월1일까지 진행된다.

이번 2022년도 학회에서는 인공지능 분야에 대한 높은 관심을 반영하듯 총 9,251편의 논문이 접수되었고, 이 중 15%인 1,349편의 논문이 채택되었다.

자세한 내용은 아래의 링크와 논문 초록을 참고해주십시오.

* 논문 관련 링크

We develop a new method of online inference for a vector of parameters estimated by the Polyak-Ruppert averaging procedure of stochastic gradient descent (SGD) algorithms. We leverage insights from time series regression in econometrics and construct asymptotically pivotal statistics via random scaling. Our approach is fully operational with online data and is rigorously underpinned by a functional central limit theorem. Our proposed inference method has a couple of key advantages over the existing methods. First, the test statistic is computed in an online fashion with only SGD iterates and the critical values can be obtained without any resampling methods, thereby allowing for efficient implementation suitable for massive online data. Second, there is no need to estimate the asymptotic variance and our inference method is shown to be robust to changes in the tuning parameters for SGD algorithms in simulation experiments with synthetic data.