How to make intelligent web-apps
summary
Builds a Django news aggregator and implements clustering with k‑means with distance threshold (canopy‑style) using set‑based Jaccard distance; tests on ~500 articles from 8 sources produce about 120 clustered items and convergence in roughly 8–10 iterations (defaults to 50). Recommends scikit‑learn, MDP, libSVM and Prediction API, shows a Naive Bayes tweet sentiment model that labels examples correctly, and notes Amazon Mechanical Turk for human labeling when accuracy demands exceed automated methods.
AI-written summary · about
chapters (11)
- 0:00Start
- 2:00Why Python
- 6:33Why We Need Intelligence in Our Application
- 9:30What Exactly Is a Machine Learning
- 10:11Google News
- 11:10Supervised and Unsupervised
- 19:31Grouping Algorithms
- 28:12Classification
- 29:46Example Sample App
- 41:24Is There an Upper Bound on the Convergence
- 45:57K-Means Clustering
related talks
37mBuild a Data-Driven Web App That Everyone Can Use
01h 17mOne workshop that data scientists don't want you to attend...
50mBuilding Intelligent Mobile Apps using Machine Learning
33mPeople who liked this talk also liked ... Building Recommendation...
30mKick-Start your Understanding of Machine Learning with Python
42mMachine learning in your hobby project: Obstacles, entry points and solutions
39mHow we Used NLP and Django to Build a...
55mBuilding a Recommendation Engine with Machine Learning Techniques