Presented By O’Reilly and Cloudera
Make Data Work
March 5–6, 2018: Training
March 6–8, 2018: Tutorials & Conference
San Jose, CA

Operationalize deep learning: How to deploy and consume your LSTM networks for predictive maintenance scenarios

Francesca Lazzeri (Microsoft), Fidan Boylu Uz (Microsoft)
11:00am–11:40am Thursday, March 8, 2018
Data engineering and architecture
 Location: LL21 C/D
Download slides (PDF)

Who is this presentation for?

  • Technical program managers, data scientists, and solution architects

Prerequisite knowledge

  • Familiarity with deep learning, Python, and predictive maintenance scenarios

What you'll learn

  • Learn how to operationalize LSTM networks to predict the remaining useful life of aircraft engines

Description

Deep learning has proven to show superior performance in certain domains such as object recognition and image classification. It has also gained popularity in domains such as finance where time series data plays an important role. Predictive maintenance, where data is collected over time to monitor the state of an asset with the goal of finding patterns to predict failures, also benefits from deep learning algorithms. Long short-term memory (LSTM) networks are especially appealing to the predictive maintenance domain due to the fact that they are very good at learning from sequences, making it possible to look back for longer periods of time to detect failure patterns.

Francesca Lazzeri and Fidan Boylu Uz explain how to operationalize LSTM networks to predict the remaining useful life of aircraft engines. They use simulated aircraft sensor values to predict when an aircraft engine will fail in the future so that maintenance can be planned in advance. Francesca and Fidan share their data science process, from data ingestion to operationalization, in a Jupiter notebook with the Keras deep learning library with Microsoft Cognitive Toolkit CNTK as a backend.

Photo of Francesca Lazzeri

Francesca Lazzeri

Microsoft

Francesca Lazzeri is a senior machine learning scientist at Microsoft on the cloud advocacy team and an expert in big data technology innovations and the applications of machine learning-based solutions to real-world problems. Her research has spanned the areas of machine learning, statistical modeling, time series econometrics and forecasting, and a range of industries—energy, oil and gas, retail, aerospace, healthcare, and professional services. Previously, she was a research fellow in business economics at Harvard Business School, where she performed statistical and econometric analysis within the technology and operations management unit. At Harvard, she worked on multiple patent, publication and social network data-driven projects to investigate and measure the impact of external knowledge networks on companies’ competitiveness and innovation. Francesca periodically teaches applied analytics and machine learning classes at universities and research institutions around the world. She’s a data science mentor for PhD and postdoc students at the Massachusetts Institute of Technology and speaker at academic and industry conferences—where she shares her knowledge and passion for AI, machine learning, and coding.

Photo of Fidan Boylu Uz

Fidan Boylu Uz

Microsoft

Fidan Boylu Uz is a senior data scientist on the algorithms and data science team at Microsoft, where she is responsible for successful delivery of end-to-end advanced analytics solutions. Fidan has 10+ years of technical experience in machine learning and business intelligence and has worked on projects in multiple domains such as predictive maintenance, fraud detection, mathematical optimization, and deep learning. She is a former professor at the University of Connecticut, where she conducted research and taught courses on machine learning theory and its business applications. She has authored a number of academic publications in the areas of machine learning and optimization. Fidan holds a PhD in decision sciences.