Autoplay
Autocomplete
Previous Lesson
Complete and Continue
Introduction to Post Deployment Data Science
Chapter 1 - Introduction
Lesson 1.1 - Why you should learn post-deployment data science (4:22)
Chapter 2 - MLOps Maturity
Lesson 2.1 - Introduction to MLOps (3:43)
Lesson 2.2 - MLOps Maturity Levels (8:14)
Chapter 3 - What is Machine Learning Monitoring?
Lesson 3.1 - Introduction to Monitoring Machine Learning Models (1:45)
Lesson 3.2 - Benefits of a Machine Learning Monitoring System (2:56)
Lesson 3.3 - Challenges of Monitoring Models in Production (3:47)
Lesson 3.4 - 91% of models degrade in production (2:45)
Chapter 4 - Overview of ML Monitoring Workflow
Lesson 4.1 - The problem with traditional monitoring (3:52)
Lesson 4.2 - The ideal monitoring workflow (3:39)
Chapter 5 - Intro to nannyML
Lesson 5.1 - What is nannyML (2:28)
Lesson 5.2 - nannyML functionality overview (3:54)
Lesson 5.3 - Ways of installing nannyML (0:23)
Lesson 5.4 - nannyML Documentation and Support
Chapter 6 - Introduction to the dataset
Lesson 6.1 - Intro to the taxi green dataset (6:14)
Lesson 6.2 - Partitioning the data and building simple regressor and classifier (2:22)
Chapter 7 - Data Requirements for Model Monitoring
Lesson 7.1 - Reference and analysis explained (0:50)
Lesson 7.2 - Data collection for effective ML monitoring (4:26)
Lesson 7.3 - How to format your data for NannyML (3:36)
Chapter 8 - Monitoring Workflow: Performance Monitoring
Lesson 8.1 - Performance Calculation (5:25)
Lesson 8.2 - Chunking (2:59)
Lesson 8.3 - Performance Estimation (8:56)
Lesson 8.4 - Business Value (3:16)
Lesson 8.5 - Deep Dive in CBPE (8:14)
Lesson 8.6 - Deep Dive in DLE (4:15)
Lesson 8.7 - Exercise: Performance monitoring for regression with NannyML (6:53)
Lesson 8.8 - Exercise: Performance monitoring for classification with NannyML (8:10)
Chapter 9 - Monitoring workflow: Root Cause Analysis
Lesson 9.1 - What is a Root Cause Analysis of a Model Failure (0:55)
Lesson 9.2 - Multivariate drift detection (1:52)
Lesson 9.3 - Exercise: Multivariate drift detection (2:03)
Lesson 9.4 - Overview of univariate drift detection methods (1:24)
Lesson 9.5 - Jensen-Shannon Distance (3:13)
Lesson 9.6 - Wasserstein Distance (2:03)
Lesson 9.7 - Hellinger Distance (1:57)
Lesson 9.8 - L-Infinity Distance (1:36)
Lesson 9.9 - Kolmogorov-Smirnov Test (1:45)
Lesson 9.10 - Chi-squared Test (2:14)
Lesson 9.11 - Prediction drift and Target drift (2:22)
Lesson 9.12 - Exercise: Univariate Drift Detection (6:14)
Lesson 9.13 - Data Quality Checks (3:26)
Lesson 9.14 - Exercise: Data Quality Checks (2:54)
Lesson 9.15 - Summary Statistics (0:43)
Lesson 9.16 - Exercise: Summary Statistics (3:27)
Lesson 9.17 - Ranking (1:05)
Lesson 9.18 - Exercise: Ranking (4:04)
Chapter 10: Monitoring Workflow: Issue Resolution
Lesson 10.1 - Strategies to fix model performance failures (0:29)
Lesson 10.2 - Retraining the Model (0:50)
Lesson 10.3 - Change Business Process (1:26)
Lesson 10.4 - Reverting back to a previous model (0:23)
Lesson 10.5 - Inducing the new drift to the old data (0:42)
Lesson 10.6 - Refactor the use case (0:30)
Lesson 10.7 - Doing nothing (0:38)
Chapter 11 - Introduction to Docker and Grafana
Lesson 11.1 - What is Docker and Grafana (2:23)
Lesson 11.2 - Installing Docker Desktop (0:40)
Chapter 12 - Putting all together with continuous monitoring
Lesson 12.1 - Overview of monitoring system in production (2:08)
Lesson 12.2 - Exercise: Grafana Analysis
Lesson 1.1 - Why you should learn post-deployment data science
Lesson content locked
If you're already enrolled,
you'll need to login
.
Enroll in Course to Unlock