Intelligent Models of the Quantitative Behavior of Microbial Systems

Abstract

Under different environmental conditions, microbial systems display complex behavioral patterns that are difficult to express quantitatively by mechanistic methods. Therefore, two alternate approaches based on different forms of intelligence have emerged. One approach uses methods of artificial intelligence (AI) such as neural networks, expert systems, and genetic algorithms to describe cellular behavior. The second methodology, which leads to the class of cybernetic models, relies on intelligence postulated to be possessed by the cells themselves. While both AI and cybernetic methods have been effective in many applications where mechanistic models are inadequate, all three methods have strengths and weaknesses. This recognition has led to the development of hybrid systems that combine two or more approaches for different aspects of a microbial system. However, the optimum design of hybrid systems still remains heuristic. The rationale and the developments from mechanistic to hybrid models are discussed here, and it is suggested that eventually a truly intelligent system should be self-evolving to maintain itself at the optimum configuration at all times.