A weekly list of interesting reads and resources.
AI / Machine Learning
- How to train your robot – Learn how to create a simple AI to turn your home lights on and off using Microsoft Azure’s learning understanding service, LUIS.
- Object-Oriented Deep Learning – The paper describes using an object-oriented neural network makes the network interpretable, with visual knowledge throughout network layers.
- The Foundations of Algorithmic Bias – I think this says it all:
“public discourse rarely throws light on the precise mechanisms by which bias actually enters algorithmic decision-making processes.”
- You Don’t Have to Be a Data Scientist to Fill This Must-Have Analytics Role – I’ve never heard of an Analytic Translator, but like most job functions in data science its a hybrid. After reading this article, I would describe an Analytics Translator as being similar to a traditional business analysis role, where you have a person that sits between the technical team and stakeholders, with the ability to translate technical-to-business value.
- Work Stealing: What Makes the Play Framework Fast – This is the best explanation of the Play Framework I’ve come across. Play Framework is a high performance, lightweight web framework for Java and Scala development.
- Five Simple Database Design Errors You Should Avoid and Lookup Table Madness – This week I revisited the topic of Lookup Tables (aka MUCK or One True Lookup Table) to determine best practices for structuring lookup data. In the end, I stay away from enums and create an individual lookup table per entity. Let me know if you have an alternate, efficient data structure for lookup data.