What's the Buzz with Machine Learning
summary
Project uses Python3 with Pandas, Bokeh, scikit‑learn and AWS to train linear regression models on Bee Informed Partnership hive weight data sampled every 15 minutes. 13 features yield 306 models tested against six hives; 25 models meet quality checks and one deployed in production to flag event windows (harvest, feeding, hive changes). Model detects farmers’ annotations but struggles with poor annotations and missing intervals which were handled later by simple rules. Overall sample coverage is 8.1%.
AI-written summary · about
chapters (10)
- 0:00Intro
- 0:32Outline
- 1:38The Three P's
- 3:16Basic Process
- 5:16Data Source
- 7:58Short Term Goals
- 16:28Good Models
- 17:38Current ML Algorithm
- 18:38Current State & What Next
- 19:33Thank You!
related talks
01h 23mTutorial: A Practical Introduction To Machine Learning
20mFINDING BAD BEES IN THE HIVE
36mDetecting Honeybee Swarms Using the Integration of OpenCV, Pandas, AI, & PyTorch
22mReal-world applications of machine learning: Beyond the hype
36mMachine Learning with Apache Beam
20mShip your Machine Learning Application #bbuzz
25mML model from an idea to production with the help...
29mMachine learning in production