Published Jun 11, 2021

Detecting Drift

Research Data Scientist Sam Ackerman delves into the intricacies of machine learning model performance and data drift detection, highlighting the importance of changepoint modeling and its challenges in ensuring model accuracy and scalability in large-scale environments.

Episode Highlights

Topics covered

Episode Highlights

Related Episodes

Change Point Detection Algorithms
Answers 383 questions
Fraud Detection in Real Time
Answers 383 questions
Machine Learning Done Wrong
Answers 383 questions
Algorithmic Detection of Fake News
Answers 383 questions
Change Point Detection in Continuous Integration Systems
Answers 383 questions
Video Anomaly Detection
Answers 383 questions
Detecting Cheating in Chess
Answers 383 questions
Detecting Fast Radio Bursts with Deep Learning
Answers 383 questions
Applying k-Nearest Neighbors to Time Series
Answers 383 questions
Detecting Terrorists with Facial Recognition?
Answers 383 questions
Predictive Models on Random Data
Answers 383 questions
Stealing Models from the Cloud
Answers 383 questions
Human Detection of Fake News
Answers 383 questions
ML Ops
Answers 383 questions
Kalman Filters
Answers 383 questions

Dexa/Kyle Polich

Detecting Drift

Topics covered

Popular Clips

Novel Scoring Function

Scalability and Resource Demands

ML Powered Analytics

Confidence and Misclassification

Solving Drift

Analyzing Model Quality

Detecting Data Drift

Fine-tuning Alpha

Detecting Drift in Handwritten Digits

Unveiling Data Insights

Detecting Data Drift

Drift and Model Performance

Unpacking the Changepoint Model

Episode Highlights

Model Performance Analysis

Detecting Data Drift

Changepoint Modeling

Related Episodes

Change Point Detection Algorithms

Fraud Detection in Real Time

Machine Learning Done Wrong

Algorithmic Detection of Fake News

Change Point Detection in Continuous Integration Systems

Video Anomaly Detection

Detecting Cheating in Chess

Detecting Fast Radio Bursts with Deep Learning

Applying k-Nearest Neighbors to Time Series

Detecting Terrorists with Facial Recognition?

Predictive Models on Random Data

Stealing Models from the Cloud

Human Detection of Fake News

ML Ops

Kalman Filters

Detecting Drift

Topics covered

Popular Clips

Episode Highlights

Model Performance AnalysisSam Ackerman explores the tools and methods used to diagnose machine learning model performance and detect areas with lower accuracy. He also examines how confidence levels in predictions can indicate potential errors or misclassifications.

Model Performance Analysis

Detecting Data DriftSam Ackerman, a Research Data Scientist at IBM, discusses the challenges of detecting data drift and its impact on machine learning models. He emphasizes the importance of understanding drift to ensure model accuracy and reliability over time.

Detecting Data Drift

Changepoint ModelingSam Ackerman explores the mechanics and challenges of changepoint models in detecting data shifts. He discusses their application in machine learning and the scalability issues faced in large-scale environments.

Changepoint Modeling

Related Episodes