Computational Intelligence
What is computational intelligence and what is the difference between it and traditional AI?
The first definition of computational intelligence was proposed by BEZDEC (Bezdek) in 1992. He believes that computational intelligence depends on the numerical data provided by the manufacturer, not on knowledge, and on the other hand, the application of artificial intelligence knowledge. Artificial neural networks should be called computational neural networks.
It may not be appropriate to classify neural networks (NN) in artificial intelligence (AI), and the Classification of Computational Intelligence (CI) can be more descriptive of the substance of the problem. Some topics in evolutionary computing, artificial life and fuzzy logic systems are also classified as computational intelligence
Bezdec give certain symbols and brief descriptions or definitions to these relevant terms. He gives an interesting ABC:
A-artificial: Artificial (non-biological), artificial
B-biological, which represents the physical + chemical + (??) = Creature's
C-computational, representing math + computer
The following figure shows the relationship between ABC and its relation to neural network (NN), pattern recognition (PR), and Intelligence (I).
Transverse: nn->pr->i (neural network-> pattern Recognition-> Intelligence)
Portrait: c->a->b (the-> symbol of the value of the-> creature)
Computational Intelligence is a low-level cognition of intelligence, and the difference between it and AI is that the level of cognition falls from the bottom to the lower levels. Middle-tier systems contain knowledge (boutique), while low-level systems do not.
When a system involves only numerical (low-level) data, contains the pattern recognition part, does not apply the knowledge of artificial intelligence, and can show:
(1) Computational adaptability;
(2) Calculation of fault tolerance;
(3) The speed of approaching people;
(4) The error rate is similar to the person,
Then the system is a computational intelligence system.
When an intelligent computing system adds a value of knowledge (fine) in a non-numeric way, it becomes an AI. Classification of computational Intelligence:
Computational intelligence mainly consists of three parts
Neural computing
Artificial neural Network algorithm
Fuzzy calculation
Fuzzy logic
Evolutionary computing
Genetic algorithm (evolutionary strategy, evolutionary planning)
Ant colony optimization algorithm
Particle swarm optimization algorithm
Immune algorithm
Characteristics of computational Intelligence:
of intelligence
Including the adaptive and self-organization of the algorithm, the algorithm does not depend on the characteristics of the problem itself, with Universal
Parallelism
The algorithm is optimized to solve the problem by group collaboration, which is suitable for large-scale parallel processing.
of robustness
The algorithm has good fault tolerance and is insensitive to initial conditions, and can find the optimal solution under different condition.