PostgreSQL and Artificial intelligence
summary
Shows a working proof of concept that trains a tiny neural network inside PostgreSQL using plpython and PL/Tcl style tensors, generating a random dataset, doing an 80/20 split for train/test, and storing weights and results in tables. With ~100–1,000 rows the model only hits ~0.22 or single-digit accuracy, demonstrating it runs but needs real volumes. Argues in‑database ML avoids CSV exports, enables continuous retraining for fraud detection, wraparound prediction and config tuning, and can leverage Postgres GPU extensions.
AI-written summary · about
chapters (10)
- 0:00Start
- 2:47Three Types of Artificial Intelligence
- 5:48Sub Fields of Artificial Intelligence
- 7:58Machine Learning
- 8:12Types of Machine Learning
- 11:49Unsupervised Learning
- 15:52Deep Learning
- 19:06The Difference between the Machine Learning and the Deep Learning
- 23:46Postgresql Have Different Languages
- 42:36Resource Usage
related talks
45mPostgres and the Artificial Intelligence
25mPostgres and the Artificial Intelligence Landscape
49mPostgres and the Artificial Intelligence Landscape
35mPostgres in Artificial Intelligence
50mFrom Transactions to Intelligence: Postgres for Open AI & Analyti
37mRound Table Discussion
53mModern PostgreSQL for AI...
27mPutting AI in Production