CLASSIFICATION OF ASTEROID SPECTRA USING A NEURAL-NETWORK

The 52-color asteroid survey (Bell et al., 1988) together with the 8-c olor asteroid survey (Zellner et al., 1985) provide a data set of aste roid spectra spanning 0.3-2.5 mum. An artificial neural network cluste rs these asteroid spectra based on their similarity to each other. We have also trained the neural network with a categorization learning ou tput layer in a supervised mode to associate the established clusters with taxonomic classes. Results of our classification agree with Thole n's classification based on the 8-color data alone. When extending the spectral range using the 52-color survey data, we find that some modi fication of the Tholen classes is indicated to produce a cleaner, self -consistent set of taxonomic classes. After supervised training using our modified classes, the network correctly classifies both the traini ng examples, and additional spectra into the correct class with an ave rage of 90% accuracy. Our classification supports the separation of th e K class from the S class, as suggested by Bell et al. (1987), based on the near-infrared spectrum. We define two end-member subclasses whi ch seem to have compositional significance within the S class: the So class, which is olivine-rich and red, and the Sp class, which is pyrox ene-rich and less red. The remaining S-class asteroids have intermedia te compositions of both olivine and pyroxene and moderately red contin ua. The network clustering suggests some additional structure within t he E-, M-, and P-class asteroids, even in the absence of albedo inform ation, which is the only discriminant between these in the Tholen clas sification. New relationships are seen between the C class and related G, B, and F classes. However, in both cases, the number of spectra is too small to interpret or determine the significance of these separat ions.