statistics – Page 6 – Andrew Fairless, Ph.D.

What I Read: explainability, survival analysis

By Andrew Fairless on November 21, 2022September 27, 2022

https://medium.com/responsibleml/survex-model-agnostic-explainability-for-survival-analysis-94444e6ce83d survex: model-agnostic explainability for survival analysisMikołaj SpytekSep 19 “… survival models… tells us what is the probability of an event not happening until a given time t…. The complexity…Continue readingWhat I Read: explainability, survival analysis

What I Read: Neural Tangent Kernel

By Andrew Fairless on November 2, 2022September 27, 2022

https://lilianweng.github.io/posts/2022-09-08-ntk/ Some Math behind Neural Tangent KernelLilian WengSeptember 8, 2022 “Neural tangent kernel… leads to great insights into why neural networks with enough width can consistently converge to a globalContinue readingWhat I Read: Neural Tangent Kernel

What I Read: Causal Inference

By Andrew Fairless on October 25, 2022September 9, 2022

https://thegradient.pub/causal-inference-connecting-data-and-reality/ Causal Inference: Connecting Data and RealityWenwen Ding04.Sep.2022 “Causal inference is a theory that describes, discriminates, and measures causal relationships, developed from statistics.”

What I Read: Comparing quantiles at scale

By Andrew Fairless on October 3, 2022August 22, 2022

https://engineering.atspotify.com/2022/03/comparing-quantiles-at-scale-in-online-a-b-testing/ Comparing quantiles at scale in online A/B-testingMarch 23, 2022Published by Mårten Schultzberg (Senior Data Scientist) and Sebastian Ankargren (Data Scientist) “…we have been able to reduce the computational complexityContinue readingWhat I Read: Comparing quantiles at scale

What I Read: Concept Drift Without Labeled Data

By Andrew Fairless on September 26, 2022August 22, 2022

https://concept-drift.fastforwardlabs.com/ Inferring Concept Drift Without Labeled DataFF22Aug 2021 “…we explore broadly applicable approaches for dealing with concept drift when labeled data is not readily accessible. We’ll start by defining whatContinue readingWhat I Read: Concept Drift Without Labeled Data

What I Read: introduction to NumPyro

By Andrew Fairless on September 19, 2022August 22, 2022

https://dfm.io/posts/intro-to-numpyro/ An astronomer’s introduction to NumPyroDan Foreman-MackeyJul 28 2022 “In this post… I’ll focus primarily on providing an introduction to NumPyro, which is a probabilistic programming library that provides anContinue readingWhat I Read: introduction to NumPyro

What I Read: Supercharging A/B Testing

By Andrew Fairless on September 15, 2022August 3, 2022

Supercharging A/B Testing at Uber Supercharging A/B Testing at UberSergey Gitlin, Krishna Puttaswamy, Luke Duncan, Deepak Bobbarjung, and Arun Babu A S PJuly 21, 2022 “…building a correct and reliableContinue readingWhat I Read: Supercharging A/B Testing

What I Read: Short Time Series

By Andrew Fairless on September 1, 2022August 3, 2022

https://minimizeregret.com/short-time-series-prior-knowledge Modeling Short Time Series with Prior KnowledgeWhat ‘Including Prior Information’ really looks like.Tim Radtke2019-04-16 “It is generally difficult to model time series when there is insuffient data to modelContinue readingWhat I Read: Short Time Series

What I Read: Anomaly Detection Metrics

By Andrew Fairless on August 25, 2022July 18, 2022

https://medium.com/@katser/a-review-of-anomaly-detection-metrics-with-a-lot-of-related-information-736d88774712 A Review of Anomaly Detection MetricsIurii KatserJul 12 “Anomaly detection… is a problem… to identify unusual patterns that do not conform to expected behavior…. we will talk about anomalyContinue readingWhat I Read: Anomaly Detection Metrics

What I Read: Estimating Model Performance

By Andrew Fairless on August 23, 2022July 18, 2022

https://pub.towardsai.net/estimating-model-performance-without-ground-truth-453b850dad9a Estimating Model Performance without Ground TruthIt’s possible, as long as you keep your probabilities calibratedMichał OleszakMay 12 “A model in production is like a baby: it needs to beContinue readingWhat I Read: Estimating Model Performance

Tag: statistics