Main Menu

search

You are here

Artificial Intelligence

[last updated: 2024-04-12]
neural nets
-----


      On This Page:
  • Algoritms
  • Computer systems
  • AI: definition
  • statistical concepts
  • Links/refs

---------------------------------------------

  • General:
    • Algorithm:
      The steps or calculation or series of calculations
      that are performed on the input data
      to produce the output.

    • Computer systems:
      take some input data,
      apply some algorithm to it,
      and get a result (output),
          perhaps an action,
          perhaps a prediction,
          perhaps an evaluation or analysis.

    • Artificial Intelligence:

        Artificial Intelligence computer systems differ from "conventional" computer systems
            in the way they get their algorithms:
          In conventional programming, the algorithms, the calculations needed to get from input to output,
          are generally "known,"
          and can therefore be explicitly programmed into the code,


          in AI systems,
          the system "discovers" or "learns" its own algorithms,
          by processing "test" data sets, for which the correct/desired outputs are also known.



      As a result, AI computer systems tend to be more complex than conventional systems.
      Their actual algorithms, to the extent that they are even known, can be significantly more complicated than standard systems.
      The data sets they can work with are also typically more complicated, in that there are more inter-relations between parameters.

    ---------------------------------

  • Building an AI system requires several steps:
    • hardware:
        standard computers
        neural nets (or simulations of them)
    • data collection
    • computation algorithm:
      • you have to start with a model
        this is a description, based on prior knowledge,
        of the general dependencies between inputs and outputs
          [from ref-1]:
          "Each [type of] model corresponds to a certain type of dependency assumption between the inputs and the output"
      • ... then your system processes the input data against that model,
        and finds a formula/algorithm that gives the desired output
    • output

---------------------------------

  • Supervised Learning:
    • regression is one type of SL.

    ---------------------------------

  • Random notes...
    ---------------------------------

  • ...
    ---------------------------------

  • Statistical concepts:
    • inference: going from specific observations to general descriptions
    • estimation: the process of learning
    • discriminant analysis: classification ...

    ---------------------------------

  • Links/Refs:

    .

    .

    eof