Embedding Layers and Why You Need to Start Using Them

We are pleased to have Bryan Mosher presenting on Embedding Layers and Why You Need to Start Using Them!

Embedding is growing in popularity among deep learning enthusiasts. It offers a better way to represent sparse data versus bag-of-words models and one-hot-encoding. The result is a dense representation that uncovers the relative meaning between features. Embeddings are learned during training, can be reused in other models and visualized using algorithms such as tSNE.

This presentation will walk through a python code-level example that uses embedding so that you walk away with an understanding of what embedding is and why you should consider using them in your deep learning modeling.

Bryan Mosher is the Chief Data Scientist at AgileThought and is responsible for the strategic direction of the data science community of practice within AgileThought and for the growth of its services. Prior to joining AgileThought, Bryan ran a machine learning consulting company, where he applied his engineering and machine learning expertise in the FinTech industry. His recent experience also includes 5 years as Chief Technology Officer at Medalogix, where he joined founder, Dan Hogan, in 2012. At Medalogix, Bryan was responsible for the architecture and development of the home health and hospice industry's first predictive modeling platform. The company achieved notoriety in its recognition by Harvard University as a health acceleration challenge award winner, where it was the focus of a Harvard Business School case study.