Bayes’ Theorem’s Importance in Biology: Animal Behavior
The main concept that governs Bayes’ Theorem is that one can modify the existing probably of an event happening based on observations. In other words, if you are given a probability that an event will occur and after some time, some parameters change. You can then predict the new probability that that event will occur given the new parameters. John M. McNamara, Richard F. Green and Ola Olsson, in their article Bayes’ Theorem and Its Applications in Animal Behavior, used the fundamental ideas from Bayes’ Theorem to analyze animal behaviors to see if animals use prior knowledge and integrate it into their daily lives. An important term that was brought up in this article is Statistical Inference, which is taking the prior probability (the original probability of an event happening) making observations and modifying the prior probability by means if Bayes’ Law and then making inferences of the event based of off the new probabilities.
There are two main ways animals use prior knowledge in their daily lives. The first of these two is adaption. Adapting to situations and surroundings is rooted in evolution. If a species has the ability to adapt to changes in its environment, as a result through natural selection it is more likely to survive and pass this trait down to the next generation. Unlike a learned skill this trait is innate and usually leaves no room for creative self-improvement. The article points out that animals that use this behavior may continue to do that action even if what they predicted turned out to be incorrect. Adaption in many ways restrict the animal’s ability to learn. Experience is on the opposite side of the spectrum and depends on the animal’s ability to learn from its environment. These are decisions that the animal makes based on its past experiences. However, the environment can also play a role in affecting the animal’s decisions.
The article gives three examples which demonstrate animals using prior knowledge to predict future outcomes in a way using a theoretical approach to Bayes’ Rule. The first example is foraging for food. While an animal is searching in a certain patch for food, it will notice how frequent it comes across food. When moving to another patch, it if looks similar to one where it found no food, it will skip over the patched which were similar to that one. In this example the animals is using it prior experience to predict its outcome if it were to search in a similar patch. The second example is choosing a mate during annual breeding season. In this example, the females choose the male mate by inspection. Every year, there is a new group of males to choose from, so the females do not know what the range of selection would be at the beginning of the season. However, they know from prior experience what to expect. The final example deals with growing in an environment with certain predatory risks. The prior probability is how the animal’s ancestors were able to survive. Now taking into account how well the current animal adapts to the environment and the current animal’s skill set as well as the frequency of predators, the prior probability can be modified to model the current situation. These three examples demonstrate how Bayes’ Theorem can be used to model animal behavior.
Article Title: Bayes’ Theorem and Its Applications in Animal Behavior
Source: McNamara, J. M., Green, R. F., & Olsson, O.. (2006). Bayes’ Theorem and Its Applications in Animal Behavior. Oikos, 112(2), 243–251. Retrieved from http://www.jstor.org/stable/3548663