The function of adult neurogenesis in the dentate gyrus is not yet completely understood, though many competing theories have attempted to explain the function of these newly -generated neurons. Most theories give adult neurogenesis a role in aiding known hippocampal/dentate gyrus functions. Other theories offer a novel role for these new cells based on their unique physiological qualities, such as their low excitability threshold. Many behavioral tests have been used to test these theories, but results have been inconsistent and often contradictory. Substantial variability in tests and protocols may be at least partially responsible for the mixed results. On the other hand, conflicting results arising from the same tests can serve as aids in elucidating the function of adult neurogenesis. Here, we offer a hypothesis that considers the cognitive nature of tasks commonly used to assess the function of adult neurogenesis, and introduce a dichotomy between tasks focused on discrimination vs. generalization. We view these two aspects as opposite ends of the continuous spectrum onto which traditional tests can be mapped. We propose that high neurogenesis favors behavioral discrimination while low adult neurogenesis favors behavioral generalization of a knowledge or rule. Since many tasks require both, the effects of neurogenesis could be cancelled out in many cases. Although speculative, we hope that our view presents an interesting and testable hypothesis of the effect of adult neurogenesis in traditional behavioral tasks. We conclude that new, carefully designed behavioral tests may be necessary to reach a final consensus on the role of adult neurogenesis in behavior., A. Pistikova, H. Brozka, A. Stuchlik., and Obsahuje bibliografii
Generalization phenornena which také plače in two different assembly neural networks are considered in the paper. Either of these two assembly networks is artificially partitioned into several subiietworks according to the number of classes that the network has to recognize. Hebb’s cissernblies are formed in the networks. One of the assembly networks is with binary connections, the other is with analog ones. Recognition abilities of the networks are compared on the task of handwritten character recognition. The third neural network of a perceptron type is considered in the paper for comparison with the previous ones. This latter network works according to the nearest-neighbor method. Computer simulation of all three neural networks was performed. Experirnents showed that the assembly network with binary connections has approximately the same recognition accuracy as the network realizing the nearest-neighbor technique.