In the Name of Whiskey
summary
Uses TensorFlow to run k‑means and a simple neural net on a homemade Scotch dataset, reporting k‑means vs neural net results and cluster centroids visualized on a map. Finds only about 86 data points, misaligned coordinates, and messy labels, so accuracy hovers near coin‑flip. Builds pipelines to featurize notes, regions and lat/long and concludes failed to predict flavor from coords without far more curated data and examples.
AI-written summary · about
chapters (33)
- 0:00Introduction
- 2:42What is Machine Learning
- 3:02Can I Hug
- 3:29Octopus
- 3:52Hugger croissant
- 4:05Whiskey
- 4:17Scotch
- 4:56Feature Vector
- 6:00Gateway Scotch
- 6:51Notes
- 7:34Tensorflow
- 9:18Code
- 11:15Map
- 11:49The Map
- 12:02Kmeans
- 12:49Iris Dataset
- 13:04Algorithm
- 14:42Docker Container
- 15:34Cluster Assignment
- 16:02Centroid
- 16:21Previous Run
- 17:25Kmeans Code
- 18:07What is a Neural Network
- 19:14The Hidden Layer
- 20:08The Algorithm
- 23:11What I Learned
- 23:42My Scottish Accent
- 24:05The Data
- 25:30The Problem
- 26:38Practice
- 27:31Data
- 28:22Resources
- 29:30Questions
related talks
34mWhisky Clustering with Apache Projects: Groovy, Commons Math, Ignite, Spark, Wayang, Beam and Flink
48mHow to make intelligent web-apps
24mThe TL;DR of EDA
30mWords in Space
43mDeveloper Accessible Machine Learning
40mTools and Tricks from a Pragmatic Data Scientist
39mData, Science: Investigating 1 Million Galaxies with Humans and TensorFlow
30mKick-Start your Understanding of Machine Learning with Python