Simplifying Machine Learning Lifecycle Management in Healthcare (hosted by Microsoft and Toronto Machine Learning Society)
At Microsoft, we believe in empowering data scientists and developers in healthcare by simplifying the tools and resources needed to build, train, and deploy ML models while fostering research collaboration. From real-time patient wait time predictions to training models on unstructured medical images, we’ve got you covered with a wide range of best practices in creating and productionizing machine learning based solutions in healthcare.
In this workshop we will showcase the creation of an entire ML lifecycle from data ingestion to model creation, deployment, and visualization for a healthcare scenario using Microsoft Azure services.
You will learn how to:
Create and orchestrate data ingestion pipelines and secure storage for training and inferencing data.
Leverage Azure Databricks and Azure Machine Learning for feature processing
Apply AutoML in Azure Machine learning for training and hyperparameter tuning in a parallelized manner.
Deploy and score models for both parallel batch and real-time prediction using Azure machine learning and streaming analytics.
Assess explainability and fairness of trained models.
Join us as we explore how to simplify the process of getting data science to production in an Azure environment while drastically cutting down the time needed to get data science to production. We’ll show how to map a healthcare business problem into an ML production pipeline through utilizing the right tools, and ultimately how to deploy ML models in production at scale to accelerate business value.
This session will also include a live demo for real-time waiting time prediction in hospitals and batch scoring for predicting length of stay for admitted patients.