— In the News —
This is a story about social psychology and data analytics. It's a shadowy tale involving big data, billionaires, and political agendas. This is a story about using data as a weapon...
Programmatic stock trading has been in wide use for awhile but increasingly, stocks are being traded by big data and AIs. Will an AI super-investor eventually emerge? And if so, then what?
The United Nations has a rich history of collecting important information but it's often locked up on paper or in PDFs. Or sometimes the data is in a useful format but it's on someone's laptop somewhere. These problems aren't surprising for an international relief agency but for an organization that's not at the forefront of technology, their solution is pretty awesome.
— Tools and Techniques —
Part 1 of 3 in a series of posts that looks at the landscape of AI products and companies that "are moving the needle." This part is focused on products for personal use. Whether you're interested in playing with new tech for yourself or you're interested in how the B2C AI industry is developing, there's a lot to explore here!
Nice list of Python libraries for working with data. Includes core libraries such as NumPy and Pandas as well as libraries for visualization, machine learning, natural language processing, data mining, and statistics.
Automated Machine Learning (AML) seeks to automate many of the repetitve tasks associated with machine learning workflows. It's not about replacing data scientists. It's about empowering them to do more. This post describes how Airbnb's Data Science and Analytics team uses AML, including insights into its effectiveness.
Great guide for approaching a Kaggle competition. Covers how to explore the data, create and engineer features, build models, and submit predictions. In case you're not familiar with it, Kaggle is a very popular Data Science platform where users can share, collaborate, and compete. If you're not already familiar with Kaggle, definitely check it out.
— Resources —
Awesome data release. This anonymized dataset contains over 3 million grocery orders from more than 200,000 Instacart users. Includes an overview of the data, interesting findings, and of course, a link to the data itself. Read this to find out what people are hungry for late at night but the real intent here is to provide worthwhile data for testing prediction models that are related to consumer purchases.