Data Science articles from Oct. 2021

DOD looks to civilian workforce to close technology gaps

Junk Algorithms

OpenAI attempts to summarize two recent KDnuggets posts

A new machine learning optimization technique

The state of undergraduate Bayesian education with recommendations

Commercial remote sensing companies “pivoting marketing efforts away from the NRO and instead focusing on direct sales to other US national security customers”

The value of “small data” approaches: transfer learning, data labeling, artificial data generation, Bayesian methods and reinforcement learning

‘Small Data’ are crucial to machine learning

The US satellite imagery industry readies for the NRO’s Electro-Optical Commercial Layer program—an open competition for satellite imagery products

NGA is planning to begin testing out the concept of using a “data lakehouse” to begin breaking down the walls between where data is managed at the agency

Why the Air Force’s First Software Chief is calling it quits

Masking use of graph neural networks

A case for holding tech companies responsible for their algorithms

CodeNet (and similar projects) are paving the way for Natural Language Coding  

The US Senate is considering bill that would force the military to introduce key performance indicators measuring how effectively it used AI in operations

The problem with p-hacking is not the “hacking,” it’s the “p” 

An Inconvenient Truth About AI: the ghost in the machine is essential (for now) 

Neural networks: structure, types, and possibilities

Is Machine Learning an Art, a Science or Something Else?