报告题目：A Stochastic Neural Network for uncertainty quantification of deep neural networks

报 告 人：曹延昭教授

所在单位：美国奥本大学

报告时间：2022年9月15日 星期四 上午8:30-10:00

报告地点：#腾讯会议 ID：866-653-995

校内联系人：邹永魁 zouyk@jlu.edu.cn

报告摘要：Uncertainty quantification (UQ) of deep neural networks (DNN) is a fundamental issue in deep learning. In our UQ for DNN framework, the DNN architecture is the neural ordinary differential equations (Neural-ODE), which formulates the evolution of potentially huge hidden layers in the DNN as a discretized ordinary differential equation (ODE) system. To characterize the randomness caused by the uncertainty of models and noises of data, we add a multiplicative Brownian motion noise to the ODE as a stochastic diffusion term, which changes the ODE to a stochastic differential equation (SDE). The deterministic DNN becomes a stochastic neural network (SNN). In the SNN, the drift parameters serve as the prediction of the network, and the stochastic diffusion governs the randomness of network output, which serves to quantify the epistemic uncertainty of deep learning. I will present results on convergence and numerical experiments for the SNN.