Dempster-Shafer Evidence Fusion
This method helps a computer combine answers from several judges. Each judge looks at the same case and gives an answer, and the method keeps support for an answer, support against an answer, and the part that is still unknown.
Truth Maintenance Systems
This method helps a computer track why it believes each fact. When a source changes or a key fact fails, the system removes the beliefs that no longer have support.
Case-Based Reasoning
Each case holds the problem, the fix, and the result. A new problem arrives, and the system looks for an old case that looks most like the new problem.
Soar Chunking and Procedural Memory
This method helps an AI turn slow problem solving into a fast, reusable skill. The system first works through a hard task step by step, and then saves the path as one rule.
Reservoir Computing
This method helps a small system learn patterns in a stream of numbers. The system trains only the last step in the process, so training stays fast and cheap.
Hyperdimensional Computing
This method turns each item into a very long code of simple values. Similar items get similar codes, and the system compares a new code with stored class codes.
Resident Tripwire (one-page concept note)
A tiny model lives on each computer and learns what normal looks like for that one machine. The model raises a flag the moment the machine starts to behave in a new way.