Chap2 Fuzzy logic and ANN: Introduction¶

The fuzzy logic(FL) is an extension of the binary logic( 二元逻辑 ).

The crucial step towards the application of fuzzy logic to technical systems: to allow variables with linguistic(qualitative, fuzzy) values, to map them onto a numerical(quatitative) range and to process them at the numberical level automatically.

The major premise of fuzzy theory is the(for a human natural) incompatibility of high complexity and high precision( 高复杂性和高精度的不可兼容性 ).

FL describes logical systems in the mathmatical sense, with the aim of implementing models representing human decision making(IF-THEN rule base)

In a fuzzy set, a so-called membership function( 隶属度函数 ) assigns a numerical value to each element to the set, which is the degree of belonging to the set.

Membership degree and probablity are different in nature and have absolutely no correlation.

Chap2b Neural Networks: Introduction¶

In associative learning, which provides a learning model for the supervised trained( 有监督训练 ) artificial NN, a new response becomes associated with a particular stimulus

• instrumental conditioning learning the relationship between a stimulus(event) and a reaction
• classical conditioning learning the relationships between two different events

Structure and connnetion strengths(weights) of an ANN determine its behavior and represent the degrees of freedom during optimization.

评论区