Scikit and the K-Nearest Neighbor Algorithm
summary
Scikit learn K‑Nearest Neighbors applied to the classic iris dataset with 150 samples and two features uses an IPython notebook to build a classifier, plot decision boundaries, and compare uniform vs distance weighting. The demo sets k=37 (and notes k=149 collapses classes), shows how KNN is a lazy, high‑cost at test time method, and publishes code on GitHub so the visual classifier and results can be reproduced.
AI-written summary · about
chapters (9)
- 0:00Start
- 7:54Theoretical Data Model for Supervised
- 9:17Mystery Fruit
- 13:08Example data points for each iris species
- 18:19What to learn?
- 18:43IPython Notebooks I Stalk
- 20:40Datascience as a Service
- 21:33How do you live?
- 23:08Theoretical data model for unsupervised learning
related talks
40mTools and Tricks from a Pragmatic Data Scientist
01h 27mLearning scikit learn - An Introduction to Machine Learning in Python
32mChoose the right algorithm to help you build better ML models
47mData Science and Machine Learning with Python
01h 08mNearest Neighbors I
42mDistribution-Specific Analysis of Nearest Neighbor Search and Classification
49mSpectral Approaches to Nearest Neighbor Search
01h 05mNearest Neighbors II