Efficient Neural Network Synthesis and Its Application in Smart Healthcare

Jun 2, 2022, 10:00 am11:30 am
Zoom Meeting See Abstract for Link
Event Description

Deep neural networks (DNNs) have become the driving force behind recent artificial intelligence research. With the help of a vast amount of training data, neural networks can perform better than traditional machine learning algorithms in many applications. However, there are still several issues in the training and deployment of DNNs that limit their use in various applications. First, an important problem with implementing a DNN is the design of its architecture. Using a trial-and-error design approach is very time-consuming and leads to sub-optimal architectures. Another issue is that modern neural networks often contain millions of parameters, whereas many applications require small inference models due to imposed resource constraints. In addition, DNNs generally need large amounts of data for training. However, this requirement is not met in many settings.

To address these problems we propose DNN synthesis frameworks to generate compact and accurate DNN architectures, optimize various hyperparameters of these models, and reduce the need for large training datasets. We first introduce a two-step neural network synthesis methodology, called DR+SCANN, that combines two complementary approaches to design compact and accurate DNNs. Then, we develop the CURIOUS DNN synthesis methodology. It uses a performance predictor to efficiently navigate the architectural search space with an evolutionary search process. To address the need for large amounts of data in training DNN models, we propose the TUTOR DNN synthesis framework. TUTOR relies on the generation, verification, and labeling of synthetic data to address this challenge.

We then apply the proposed DNN synthesis frameworks to smart healthcare. The emergence of wearable medical sensors (WMSs) alongside efficient DNN models points to a promising solution for the problem of disease diagnosis on edge devices. In this area, we first propose a framework for repeated large-scale testing of SARS-CoV-2/COVID-19, called CovidDeep. CovidDeep combines efficient DNNs with commercially available WMSs for pervasive testing of the virus and the resultant disease. We also propose a framework for mental health disorder diagnosis, called MHDeep. MHDeep utilizes efficient DNN models and data obtained from sensors integrated in a smartwatch and smartphone to diagnose three mental health disorders: schizoaffective, major depressive, and bipolar.

Zoom Meeting  https://princeton.zoom.us/j/6536189535