W. C. Morris, G. W. Cottrell, and J. L. Elman. A connectionist simulation of the empirical acquisition of grammatical relations. In Hybrid Neural Systems (this volume). Springer-Verlag, 2000.  Home/Search   Document Details and Download   Summary   Related Articles   Check

.... on parallel neural and symbolic learning, which includes using (1) two separate neural symbolic algorithms applied simultaneously [78] 2) two separate algorithms applied in succession, 3) integrated neural symbolic learning [80, 35] and (4) purely neural learning of symbolic knowledge, e.g. [46, 51]. The issues described above are important for making progress in theories and applications of hybrid systems. Currently, there is not yet a theory of hybrid systems . There has been some preliminary early work towards a theoretical framework for neural symbolic representations, but to date ....

.... that language is the quintessential feature of human intelligence [85] While certain learning and architectures in humans may be innate, most researchers in neural networks argue for the importance of development and environment during language learning [87, 94] For instance, it was argued [51] that syntax is not innate and that it is a process rather than representation, and abstract categories, like subject, can be learned bottom up. The dynamics of learning natural language is also important for designing parsers using techniques like SRN and RAAM. SARDSRN and SARDRAAM were ....

W. C. Morris, G. W. Cottrell, and J. L. Elman. A connectionist simulation of the empirical acquisition of grammatical relations. In Hybrid Neural Systems (this volume). Springer-Verlag, 2000.

.... on parallel neural and symbolic learning, which includes using (1) two separate neural symbolic algorithms applied simultaneously [78] 2) two separate algorithms applied in succession, 3) integrated neural symbolic learning [80, 35] and (4) purely neural learning of symbolic knowledge, e.g. [46, 51]. The issues described above are important for making progress in theories and applications of hybrid systems. Currently, there is not yet a theory of hybrid systems . There has been some preliminary early work towards a theoretical framework for neural symbolic representations, but to date ....

.... that language is the quintessential feature of human intelligence [85] While certain learning and architectures in humans may be innate, most researchers in neural networks argue for the importance of development and environment during language learning [87, 94] For instance, it was argued [51] that syntax is not innate and that it is a process rather than representation, and abstract categories, like subject, can be learned bottom up. The dynamics of learning natural language is also important for designing parsers using techniques like SRN and RAAM. SARDSRN and SARDRAAM were ....

W. C. Morris, G. W. Cottrell, and J. L. Elman. A connectionist simulation of the empirical acquisition of grammatical relations. In Hybrid Neural Systems (this volume). Springer-Verlag, 2000.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC