O'Reilly on Data

MARCH 20, 2019

The objective here is to brainstorm on potential security vulnerabilities and defenses in the context of popular, traditional predictive modeling systems, such as linear and tree-based models trained on static data sets. If an attacker can receive many predictions from your model API or other endpoint (website, app, etc.),

Modeling 219

Modeling Machine Learning Predictive Modeling Consulting 219

O'Reilly on Data

SEPTEMBER 19, 2023

While data scientists were no longer handling Hadoop-sized workloads, they were trying to build predictive models on a different kind of “large” dataset: so-called “unstructured data.” ” There’s as much Keras, TensorFlow, and Torch today as there was Hadoop back in 2010-2012. And it was good.

Machine Learning 243

Machine Learning Testing Modeling Cost-Benefit 243

The Unofficial Google Data Science Blog

MARCH 31, 2016

These predictive posterior distributions have many uses such as in multi-armed bandit problems. Thompson sampling is an "explore/exploit" algorithm which was designed for Bayesian models, and its strategy is to randomly select a choice (called an "arm") with probability proportional to the posterior of that choice being the best option.

Modeling 40

Modeling Statistics Advertising Testing 40