Prediction of Modes of Toxic Action and toxicity of phenols with feature selection algorithms coupled with fuzzy ARTMAP
Descriptors suitable for the discrimination of four modes of toxic action (MOA) of 221 phenols with respect to the ciliate Tetrahymena pyriformis were selected by using SOM-dissimilarity measures and other well-known feature selection techniques. The performance of these methods and of ensembles of them to predict MOA classes and the toxicity log[1/IGC(50)] (mmol/ L) for each action mechanism when coupled with a fuzzy-ARTMAP neural network were assessed.
Rallo, R.,Espinosa, G., and F. Giralt. Prediction of Modes of Toxic Action and toxicity of phenols with feature selection algorithms coupled with fuzzy ARTMAP Proceedings of the Joint Meeting on Medicinal Chemistry, Vienna, Eds. P. Ettmayer and G. Ecker, Medimont SRL, F620C0176, 27-33. (ISBN 88-7587-163-9)
A predictive Fuzzy ARTMAP neural system and two hybrid networks, each combining a dynamic unsupervised classifier with a different kind of supervised mechanism, were applied to develop virtual sensor systems capable of inferring the properties of manufactured products from real process variables. A new method to construct dynamically the unsupervised layer was developed. A sensitivity analysis was carried out by means of self-organizing maps to select the most relevant process features and to reduce the number of input variables into the model. The prediction of the melt index (MI) or quality of six different LDPE grades produced in a tubular reactor was taken as a case study. The MI inferred from the most relevant process variables measured at the beginning of the process cycle deviated 5% from on-line MI values for single grade neural sensors and 7% for composite neural models valid for all grades simultaneously.
Rallo R, Ferré-Giné J, Arenas A, Giralt F (2002) Neural Virtual Sensor for the Inferential Prediction of Product Quality from Process Variables. Computers and Chemical Engineering (26) 12, 1735-54
An artificial neural network, based on fuzzy ARTMAP, that is capable of learning the basic nonlinear dynamics of a turbulent velocity field is presented. The neural system is capable of generating a detailed multipoint time record with the same structural characteristics and basic statistics as those of the original instantaneous velocity field used for training. The good performance of the proposed architecture is demonstrated by the generation of synthetic two-dimensional velocity data at eight different positions along the homogeneous (spanwise) direction in the far region (x/D=420) of a turbulent wake flow generated behind a cylinder at Re=1 200. The analysis of the synthetic velocity field, carried out with spectral techniques, POD and pattern recognition, reveals that the proposed neural system is capable of capturing the highly nonlinear dynamics of free turbulence and of reproducing the sequence of individual classes of relevant events present in turbulent wake flows. The trained neural system also yields patterns of the coherent structures embedded in the flow when presented with input data containing partial information of the instantaneous velocity maps of these events. In this way, the neural network is used as an expert system that helps in the structural interpretation of turbulence in a wake flow. (C) 2000 American Institute of Physics.
Giralt F, Arenas A, Ferre-Giné J, Rallo R, Kopp GA (2000) The simulation and Interpretation of free turbulence with a cognitive neural system. Physics of Fluids, 12, 1826
- DOI: 10.1063/1.870430
An automatic procedure based on of the Fuzzy ARTMAP neural network is applied to classify the structure embedded in two-component velocity signals measured in a turbulent wake behind a circular cylinder. A small part of the velocity field in the horizontal plane of the wake recorded at two downstream positions x/D = 30 and 150 was pre-processed to extract a set of relevant patterns from the data in order to train the network. The complete data files were tested with the trained net, obtaining nine different structures: clockwise and anticlockwise eddies, sinks, sources, four types of saddle points and jet-like motions. Comparison of the number of classes and patterns belonging to the same category at at x/D = 30 and 150 shows that the number of structures present in the wake increase with downstream position, i.e. with the development of turbulence. The jets present in the near wake appear in this preliminary analysis to be linked to the formation of double rollers in the far wake.
Ferre-Giné J, Rallo R, Arenas A, Giralt F (1997) Extraction of structures from turbulent signals. Artificial Intelligence in Engineering, 11, 413-419.
- DOI: 10.1016/S0954-1810(97)00003-4
An implementation of a Fuzzy Artmap neural network is used to detect and to identify (recognise) structures (patterns) embedded in the velocity field of a turbulent wake behind a circular cylinder. The net is trained to recognise both clockwise and anticlockwise eddies present in the u and v velocity fields at 420 diameters downstream of the cylinder that generates the wake, using a pre-processed part of the recorded velocity data. The phase relationship that exists between the angles of the velocity vectors of an eddy pattern is used to reduce the number of classes contained in the data, before the start of the training procedure. The net was made stricter by increasing the vigilance parameter within the interval [0.90, 0.95] and a set of net-weights were obtained for each value. Full data files were scanned with the net classifying patterns according to their phase characteristics. The net classifies about 27% of the recorded signals as eddy motions, with the strictest vigilance parameter and without the need to impose external initial templates. Spanwise distances (homogeneous direction of the flow) within the centres of the eddies identified suggest that they form pairs of counter-rotating vortices (double rollers). The number of patterns selected with Fuzzy Artmap is lower than that reported for template matching because the net classifies eddies according to the recirculating pattern present at the core or central region, while template matching extends the region over which correlation between data and template is performed. In both cases, the topology of educed patterns is in agreement.
Ferre-Giné J, Rallo R, Arenas A, Giralt F (1996) Identification of coherent structures in turbulent shear flows with a fuzzy ARTMAP neural network. International Journal of neural Systems, 7, 559-568.
- DOI: 10.1142/S0129065796000555