# tensorflow-embedding-feature-column-with-value

Using this TF tutorial, Classify structured data with feature columns, I was interested in this Embedding columns feature.

breed1_embedding = feature_column.embedding_column(breed1, dimension=8)


In this case, the Breed1 column which has 166 unique values can be collapsed into 8 dimensions. Here, the Breed1 column doesnt have a numeric value attached to it. However, say we have hypthothetical example like below with column Breed1_value added in. How can we still use this embedding columns technique ?