Abstract: Organizations are increasingly turning to Deep Learning models for forecasting and planning. As business forecasting becomes data-driven, model selection and evaluation become more complex: data scientists have many new tools at their disposal. Among more recently developed tools, Deep Learning models are becoming increasingly popular. This popularity often stems from success in practical applications: they have shown significant promise in their ability to handle multiple variables, uncover hidden patterns, and produce accurate forecasts.
However, these models are complex and rife with implementation pitfalls. Since the approach often seems like a 'black box' to the science team, it can be especially hard to deal with for both technical and nontechnical managers. With many moving pieces and potentially costly implementation, Deep Learning models can be more difficult to implement than traditional data-driven approaches. They can easily lead to forecasts that are unstable and unreliable, or to unacceptably high resource utilization. Finding the right balance between improved performance and increased complexity is a major challenge facing organizations seeking to capitalize on the potential of Deep Learning for forecasting.
This talk presents forecasting applications of Deep Learning approaches in an accessible, conceptual framework appropriate for both technical and nontechnical managers. Focusing on the intuition behind various approaches, we will explore how managers can tackle highly complex models by asking the right questions and using familiar model evaluation tools.
The talk is designed for business leaders, data science managers, and decision makers seeking to understand how Deep Learning approaches can be leveraged to improve forecasting and planning. We will discuss examples, explore some of the methodologies available, and address effective implementation.
Attendees will leave equipped with the tools to:
- Identify types of forecasting applications that can benefit from deep learning
- Broadly understand deep learning approaches relevant to forecasting
- Understand pitfalls related to deep learning approaches, and why simpler models may work better
- Evaluate the results of a forecasting program
Bio: Javed is an economist and data scientist with experience in banking, finance, forecasting, risk management, consulting, policy, and behavioral economics. He has led development of analytic applications for large organizations including Amazon and the Federal Reserve Board of Governors, and served as a researcher with the Office of Financial Research (U.S. Treasury). He holds a PhD in financial economics and MA in statistics from U.C. Berkeley, as well as undergraduate degrees in operations management and systems engineering from the University of Pennsylvania. Currently, Javed is a Senior Data Scientist on the Corporate Training team at Metis, where he works with companies to upskill their staff in data science and analytics.