Abstract.
In this article, we define such key concepts as sense entropy, sense energy, sense efficiency coefficient (SEC). These metrics are critical to determining and monitoring the performance of any real* AI implementation.
We give a description of the basic non-scalar tools for building real artificial intelligence with the ability to adapt to a variety of conditions of its habitat.
1. Introduction
In all natural sciences such as physics, chemistry, biology, etc., the concept of entropy is used to determine the measure of quality or uselessness of the processes under consideration. For example, in thermodynamics, the basic laws (beginnings) are formulated using the properties of entropy. This indicator is one of the key indicators in the description of any physical system of medium and above average complexity.
Below we provide a description of the sense entropy on the example of the concept of the Sense Theory [1,2,3,4,5,6,7], developed and described in seven basic articles earlier.
We also define the law of conservation of sense (conservation of sense) for any heterogeneous open sense space.
Want to read more? Please go to pdf-format file >>>