Virtual Fall Summit Encore 2023-Overview of Current AI/ML Research Methods in Occupational Health

Overview of Current AI/ML Research Methods in Occupational Health - Dr. Darabi

Pdf Summary

This document is a presentation about the current research methods in the field of artificial intelligence (AI) and machine learning (ML) in occupational health. The presentation covers various topics including the definition of AI systems, different types of AI problems, the REDECA framework for risk detection and control in accidents, and the development and evaluation of ML models. It also discusses the myths and misconceptions about AI systems, the increasing trend of AI in published articles, and the ethical considerations in AI systems.<br /><br />The presentation begins by introducing the motivation behind using AI and ML in occupational health and dispelling some common myths about AI systems. It then explains the concept of an AI system and its components, using a construction example to illustrate the system loop. The different types of AI problems are discussed, including real-time monitoring and control of safety systems, planning and decision making, educational AI, and fully automated systems.<br /><br />The REDECA framework is presented as a way to anticipate and control risk in occupational safety through the integration of AI technologies. The steps involved in developing an AI system are explained, including problem definition, data availability, machine learning task, machine learning development, workplace testing, and workplace installation. Each step is described in detail, with specific roles and considerations for each.<br /><br />The process of developing ML models is further explained, including data collection, cleaning, preprocessing, modeling, and deployment. Model evaluation methods such as confusion matrix, accuracy, and AUC-ROC curve are discussed. The presentation also addresses the ethical concerns related to AI, such as privacy and surveillance, bias and discrimination, and lack of human judgment. Tips for protecting privacy and mitigating these concerns are provided.<br /><br />Overall, this presentation provides an overview of the current research methods and considerations in the field of AI and ML in occupational health, highlighting the importance of ethical practices and careful evaluation of AI systems.

Keywords

research methods

artificial intelligence

machine learning

occupational health

AI systems

ML models

REDECA framework

real-time monitoring

data availability

ethical considerations