ConceptAI & ML
Active learning
A machine learning technique where the model is allowed to choose the data from which it learns, often focusing on instances where it is most uncertain.
Know first
Named as a prerequisite by the lectures that define this. The number is how many say so. A grey one has no page yet.
Machine learning3Data collection1Data curation1Machine learning classification1Machine learning models1Model training1Optimization1Prediction models1
Opens up
Concepts that name this one as a prerequisite.
Lecturers
7Each in their own terms, from the easiest level.
Fred Feldon
An approach to instruction that involves students in doing things and thinking about the things they are doing.Andrej Karpathy
A process of querying the vehicle fleet to collect specific edge-case data based on model uncertainty or triggers.Yaser Abu-Mostafa
A learning paradigm where instead of someone giving you a fixed data set, you query the value at a particular point or provide inputs and ask for outputs/rewards.Michael Sluydts
Using machine learning iteratively while data is being produced to guide the generation of new data.Steve Brunton
An optimization problem where a model goes to acquire more data when optimizing into a region it has never seen before.Lorenzo Zanisi
A technique used to reduce data generation cost by selecting the most informative training points.
Lectures that assume it
2Connections
15The lectures that define it or assume it, and the concepts next to it.
Concepts6
Lectures9
- Security and Privacy Risks of Browser Extensions
- Applications of Optimization in Machine Learning and Engineering
- AI for Fusion Energy: Accelerating STEP Tokamak Development
- Scalability in Autonomous Driving (CVPR 2020 Keynote)
- Epilogue: The Map of Machine Learning
- Physics, AI, and the Route to New Materials