Jason Brownlee

Stacking Ensemble With Dropout Regularization I was thinking about stacking ensembles (stacked generalization) in the sauna. Stacked ensembles overfit, so we need to regularize. Generally, we use cross-validation to ensure that the meta model is ...

Stacking Ensemble With Dropout Regularization
jasonbrownlee.me/blog/posts/s...

23.01.2025 22:29 — 👍 1    🔁 0    💬 0    📌 0

GitHub - Jason2Brownlee/Awesome-AutoML-Books: Awesome AutoML Books: Curated list of books on Automated Machine Learning Awesome AutoML Books: Curated list of books on Automated Machine Learning - Jason2Brownlee/Awesome-AutoML-Books

Awesome AutoML Books
A curated list of books for engineers on development with Automated Machine Learning (#AutoML).
github.com/Jason2Brownl...

30.12.2024 22:06 — 👍 1    🔁 0    💬 0    📌 0

GitHub - Jason2Brownlee/awesome-llm-books: Awesome LLM Books: Curated list of books on Large Language Models Awesome LLM Books: Curated list of books on Large Language Models - Jason2Brownlee/awesome-llm-books

Awesome LLM Books
This is a curated list of books for engineers on development with Large Language Models (LLMs)
github.com/Jason2Brownl...

26.12.2024 22:10 — 👍 3    🔁 0    💬 0    📌 0

Is there evidence that model performance on train and test sets have the same distributions?

Use statistical tests to confirm general model performance distributions are equivalent.

Check Model Performance Distributions:
datasciencediagnostics.com/diagnostics/...

11.12.2024 17:32 — 👍 0    🔁 0    💬 0    📌 0

Is there evidence that your train and test sets have the same distributions?

Use statistical tests to confirm that numerical and categorical distributions are equivalent.

Train/Test Data Distributions:
datasciencediagnostics.com/diagnostics/...

10.12.2024 20:44 — 👍 0    🔁 0    💬 0    📌 0

Is there evidence that the Performance Gap is real or just statistical noise?

Carefully quantify the difference between train and test set performance.

Quantify the Performance Gap:
datasciencediagnostics.com/diagnostics/...

09.12.2024 18:45 — 👍 0    🔁 0    💬 0    📌 0

Data Science Diagnostics

Data Science Diagnostics
Helpful checks for data scientists with urgent problems
DataScienceDiagnostics.com

#DataScience #MachineLearning

08.12.2024 21:57 — 👍 1    🔁 0    💬 0    📌 0

Are you sure your train/test split percentage is well chosen?

Common split percentages are just heuristics, it is better to know how your data/model behaves under different split scenarios.

Perform a split-size sensitivity analysis:
datasciencediagnostics.com/diagnostics/...

08.12.2024 21:46 — 👍 0    🔁 0    💬 0    📌 0

GitHub - Jason2Brownlee/DataScienceDiagnosticChecklist: Data Science Diagnostic Checklist Data Science Diagnostic Checklist. Contribute to Jason2Brownlee/DataScienceDiagnosticChecklist development by creating an account on GitHub.

Data Science Diagnostic Checklist
(from 10+ years of consulting)
github.com/Jason2Brownl...

03.12.2024 18:34 — 👍 2    🔁 0    💬 0    📌 0

Use code BLACKFRIDAY for 30% off the "Python Concurrency Boxed Set": superfastpython.com/python-jump-...

29.11.2024 20:46 — 👍 0    🔁 0    💬 0    📌 0

XGBoost is all you need!

XGBoost is all you need: XGBoosting.com

25.11.2024 02:48 — 👍 1    🔁 0    💬 0    📌 0