There are many features in most conventional software products that we consider intelligent: A browser that remembers your settings, a player that resumes exactly where you paused or stopped watching a movie and so on. These are however, not cases of intelligent software but rather intelligent coding.
Research in the area of artificial intelligence promise what could be considered a clue to the holy grail of computing: Software that can solve a problem with no human intervention.
Several approaches have been used and each model's strengths and weaknesses noted, though AI is still in its formative stages( That says something considering its a branch of computer science) and an area of active research(To most of you that might read as fun)
This is a log of my foray in AI, and I have to admit I am impressed by the work done in these fields of AI
Neural Networks
Simulates how the brain works and are trained to solve a certain problem domain.
Useful in forecasting, pattern recognition and data mining
Data mining
Involves finding useful relationships in large volumes of data, but unlike in a search engine where the user knows exactly what they want, the process should be software guided, to find relations a human could have overlooked
Case-Based Reasoning
Simulates how a human being solves problems based on similar past episodes(cases).
Involves retrieval of retained cases, reuse to solve current problem,revising the retrieved cases so as to solve the new problem and retaining the problem and its solution for future retrieval
[more to come later]
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment