Embeddings & Estimators

• Embeddings
• Custom Estimators

Embeddings

An embedding of a vector is another vector in a smaller dimensional space

• Manage sparse data
• Make machine learning models that use sparse data consume less memory and train faster
• Reduce dimensionality
• Increase model generalization
• Cluster observations

https://www.toptal.com/machine-learning/embeddings-in-machine-learning

Embeddings  |  Machine Learning  |  Google for Developers

Summary of Embeddings

Recommendations

https://www.learndatasci.com/tutorials/building-recommendation-engine-locality-sensitive-hashing-lsh-python

Data Driven Embeddings

Sparse Tensors

Train an Embedding

Similarity Property

Custom Estimator

• Go beyond canned estimators
• Write a custom estimator
• Gain control over model functions
• Incorporate Keras models into Estimator

Model Function

• Keras is a high-level deep neural network library that support multiple backends