Abstract: Welcome to deep learning building blocks. This is an intermediate tutorial on deep learning that focuses on how to design neural networks for various data types.
This tutorial progresses by introducing different types of data (often data that is hard for traditional ML to take advantage of). We then present neural network designs that typically work well with that type of data.
We will have three sections each corresponding to a different type of data:
1) Tabular Data: Tabular data is the data that you most often see. It is data that you can cleanly write in a table. It has a set number of rows and columns. We will go over the core NN design to tackle this type of data.
2) Categorical Data: Old ML algs can only treat each category as completely separate entities, whereas deep learning with the use of embeddings, can capture the similarities of some categories with others.
3) Variable Length Features: Most ML algs struggle to use data with variable length features, and instead use aggregations of these features. NNs can be structured to use this type of data in a better way.
Bio: Nathaniel earned his AB/SM in Computer Science from Harvard. He previously worked as a Quant and Trader at Jane Street and Goldman Sachs before transitioning into the pure tech industry. Nathaniel worked as a Data Scientist at Facebook, a Product Manager at Microsoft and a Software Engineer at Google before joining Vicarious. He is an avid reader and learner. He teaches part time at General Assembly and is developing open source teaching material for data science, machine learning, and web development.