C. Cleva et al., ADVANTAGES OF A HIERARCHICAL SYSTEM OF NEURAL-NETWORKS FOR THE INTERPRETATION OF INFRARED-SPECTRA IN STRUCTURE DETERMINATION, Analytica chimica acta, 348(1-3), 1997, pp. 255-265
A hierarchical system of small feed forward neural-networks is used to
extract structural information from infrared spectra. The top-level n
etwork gives a rough classification in five non-exclusive classes: com
pounds containing carbonyl, hydroxyl, amino groups, aromatic compounds
and ethylenic compounds. For each class, a dedicated network is desig
ned to identify more specific structural features. Depending upon thos
e structural features, the hierarchy might be extended to deeper level
s. Specialised networks are activated in a cascade-like effect by the
output of the upper-level networks. The training of each specialist ne
twork is performed using learning and test sets made of compounds iden
tified by the upper level networks as belonging to this class. Thanks
to this approach and to the optimisation of decision thresholds, the q
uality of the responses is excellent, and compounds wrongly classified
by one network do not lead automatically to other errors. One major a
dvantage of this approach is the limited size of each network involved
. Networks with few outputs are easier to optimise, and their performa
nce is better than that of larger networks. Moreover linking the respo
nse sets from the different refinement levels allows improvement of re
sponse quality and in some cases inference of other structural feature
s by combination of responses. Hierarchical neural-network systems are
well suited for the interpretation of infrared spectra. They perform
very well, and the different refinement levels of information permit g
reat flexibility in the ways they may be used. The modular organisatio
n allows modification of some parts of the system without damaging the
whole hierarchy.