Information Technology Reference
In-Depth Information
Review of various applications of the ANN
methodology
Table 5.2
Application
ANN used Reference
Preformulation studies
Determination of the physicochemical properties
(hydration characteristics, glass transition
temperature, and rheological properties) of
amorphous polymers. ANN was trained with sets
of experimental data consisting of different
polymer blends with known water-uptake profi les,
glass transition temperatures and viscosity
values.
MLP
Ebube
et al., 2000
Analysis of the crystal purity of mebendazole raw
material and its stability in a suspension
formulation. A developed ANN model confi rmed
that the characteristic absorptions in the IR
spectral region are directly proportional to the
ratio of different mebendazole crystal forms
presented in the samples.
GRNN
Agatonovic-
Kustrin
et al., 2008
Prediction of the drug intrinsic solubility. Molecular
descriptors were used as inputs for the network
training, whereas the logarithm of reciprocal
intrinsic solubility value was used as the output.
MLP
Louis et al.,
2010
Formulation studies
Formulation of ketoprofen hydrogel. The amount
of ethanol and absorption enhancer were used as
input parameters for network training, outputs
were the rate of ketoprofen penetration, lag time,
and total irritation score.
MLP
Takahara
et al., 1997
Formulation of immediate release tablet
formulation. Tablet composition, dwell time, and
compression force were used for the network
training, whereas network outputs were tablet
tensile strength, disintegration time, friability,
capping, and dissolution rate.
MLP
Shao et al.,
2006
Prediction of the phase behavior of 4 component
systems containing oil, water, and 2 surfactants.
Inputs used for the network training were
percentages of oil and water and HLB of the
surfactant blend, whilst the outputs were
oil-in-water emulsion, water-in-oil emulsion,
microemulsion, and liquid crystal containing
regions.
MLP
Alany et al.,
1999
Search WWH ::
Custom Search