Type: | Package |
Title: | Interaction Prediction Between Groundwater and Quarry Extension Using Discrete Choice Models and Artificial Neural Networks |
Version: | 1.0.0 |
Date: | 2016-11-23 |
Author: | Johan Barthelemy [aut, cre], Timoteo Carletti [aut], Louise Collier [aut], Vincent Hallet [aut], Marie Moriame [aut], Annick Sartenaer [aut] |
Maintainer: | Johan Barthelemy <johan@uow.edu.au> |
Description: | An implementation of two interaction indices between extractive activity and groundwater resources based on hazard and vulnerability parameters used in the assessment of natural hazards. One index is based on a discrete choice model and the other is relying on an artificial neural network. |
LazyData: | true |
Depends: | R(≥ 2.10.0), neuralnet |
License: | GPL-2 |
URL: | https://github.com/jojo-/quarrint |
NeedsCompilation: | no |
Packaged: | 2016-11-24 05:35:55 UTC; johan |
Repository: | CRAN |
Date/Publication: | 2016-11-24 11:31:13 |
Interaction Prediction Between Groundwater and Quarry Extension Using Discrete Choice Models and Artificial Neural Networks
Description
An implementation of two interaction indices between extractive activity and groundwater resources based on hazard and vulnerability parameters used in the assessment of natural hazards. One index is based on a discrete choice model and the other is relying on an artificial neural network.
Details
Package: | quarrint |
Type: | Package |
Version: | 1.0.0 |
Date: | 2016-11-23 |
Depends: | R(>= 2.10.0), neuralnet |
License: | GPL-2 |
This package provides two interactions indices between quarries (extractive activity) and groundwater ressources using two different methodologies, namely the discrete choice models and artificial neural networks. The design of those indices is fully detailed in Barthelemy et al. (2016).
The quarries and the groudwater ressources are described by 6 parameters, each classified into 4 modalities. These parameters are grouped into 2 distinct categories:
-
the geological, hydrogeological and piezometric contexts defining the hazard that a quarry represents;
-
the relative position of the quarry and the water catchments, the production of the catchments and the potential quality of the groundwater characterising the vulnerability of the groundwater resources.
Each of the resulting 3327 physically feasible combinations of these parameters
(out of a theoretical number of 4 6 = 4096 possible combinations) determines one
particular quarry site type. These feasible combination are provided in the
data frame quarries
.
Depending on the values of the parameters, the interaction index can then be low, medium, high or very high. The interaction level can then be used to inform a quarry operator on the required level of investigation before considering any extension of the quarry.
The method compute.interaction
provides an interface to compute the 2 interaction indices. It takes as an input
an object of type quarry
that can be constructed with the method
quarry
.
The discrete choice-based and the neural network-based indices can be
respectively be computed with the functions
compute.dc
and
compute.ann
also taking as an input an
object of type quarry
.
The package also includes the function
train.ann
to allow the training of custom
artificial neural network that can be used with the functions
compute.ann
and
compute.interaction
.
Finally an auxillary function int.in.range
is also provided to determine if a given integer is within a specified range.
Author(s)
Johan Barthelemy, Timoteo Carletti, Louise Collier, Vincent Hallet, Marie Moriame, M. and Annick Sartenaer.
Maintainer: Johan Barthelemy johan@uow.edu.au.
References
Barthelemy, J., Carletti, T., Collier L., Hallet, V., Moriame, M., Sartenaer, A. (2016) Interaction prediction between groundwater and quarry extension using discrete choice models and artificial neural networks Environmental Earth Sciences (in press)
Collier, L., Barthelemy, J., Carletti, T., Moriame, M., Sartenaer, A., Hallet, H. (2015) Calculation of an Interaction Index between the Extractive Activity and Groundwater Resources Energy Procedia 76, 412-420
See Also
neuralnet
for training and using artifical
neural network and BIOGEME to estimate discrete choice models
(http://biogeme.epfl.ch/home.html).
Examples
# creating a quarry
q <- quarry(geological.context = 2, hydrogeological.context = 4,
piezometric.context = 1, quarry.position = 4,
production.catchment = 4, quality.catchment = 3)
print(q)
# computing the interaction index
inter.idx <- compute.interaction(q)
print(inter.idx)
Coerce a Quarry to a Data Frame
Description
Coerce an object of type quarry
to a data.frame
. The attributes
to be stored in the data frame can be specified (default: keeping every
attributes).
Usage
## S3 method for class 'quarry'
as.data.frame(x, row.names = NULL, optional = NULL,
attr = "all", ...)
Arguments
x |
An object of type |
row.names |
Not used. |
optional |
Not used. |
attr |
A vector listing the dummy variables to be retained in the data frame. |
... |
Further arguments passed to or from other methods. |
Value
A data frame with the columns listed in attr
.
Note
The data frame will only stores the dummy variables of the quarry.
Author(s)
Johan Barthelemy.
Maintainer: Johan Barthelemy johan@uow.edu.au.
See Also
as.data.frame
for the generic method.
Examples
# creating a quarry
q <- quarry(geological.context = 2, hydrogeological.context = 4,
piezometric.context = 1, quarry.position = 4,
production.catchment = 4, quality.catchment = 3)
# coercing to a data frame
q.df <- as.data.frame(q)
# coercing to a data frame, keeping only "H1", "Z1" and "C2"
q.df.subset <- as.data.frame(q, attr = c("H1","Z1","C2"))
Neural Network-based Interaction Index for a Quarry
Description
Given an object of type quarry, a neural network computes the interaction index (low, medium, high or very high).
Usage
## S3 method for class 'quarry'
compute.ann(x, ann = NULL, rep = 1, ...)
Arguments
x |
A quarry object. |
ann |
The neural network used to estimate the interaction index. By default, if
set to |
rep |
The repetition of ann to be used. |
... |
Further arguments passed to or from other methods. For instance, see
|
Details
The neural network provided with the package has a feed-forward design and has a hidden layer of 7 nodes. It takes as an input a quarry constructed with the function "quarry" using the default parameters. This neural network is fully detailed in "Interaction prediction between groundwater and quarry extension using discrete choice models and artificial neural networks" (Barthelemy et al., 2016).
It is possible to use another neural network that has been trained with the
function train.ann
.
Value
A list whose elements are:
low |
The output of the ann for a low interaction level. |
medium |
The output of the ann for a medium interaction level. |
high |
The output of the ann for a high interaction level. |
very.high |
The output of the ann for a very high interaciton level. |
idx |
a string with the level of interaction ( |
Note
The quarry x
must have been created by the
quarry
. It can accept custom ranges for the
parameters values but they must be consistent with what has been used to train
the neural network ann
.
Author(s)
Johan Barthelemy.
Maintainer: Johan Barthelemy johan@uow.edu.au.
References
Barthelemy, J., Carletti, T., Collier L., Hallet, H., Moriame, M., Sartenaer, A. (2016) Interaction prediction between groundwater and quarry extension using discrete choice models and artificial neural networks Environmental Earth Sciences (in press)
Collier, L., Barthelemy, J., Carletti, T., Moriame, M., Sartenaer, A., Hallet, H. (2015) Calculation of an Interaction Index between the Extractive Activity and Groundwater Resources Energy Procedia 76, 412-420
Krieselm, D. (2007) A Brief Introduction to Neural Networks. On-line available at http://www.dkriesel.com
Ripley, B. (1996) Pattern recognition and neural networks Cambridge university press
See Also
compute.dc
to compute an interaction index
based on a discrete choice model and
compute.interaction
to predict the
interaction between between the quarry and the groundwater using both the
discrete choice-based index and the neural network-based index.
train.ann
to train a neural network and use it
as an input for this function.
compute
and
neuralnet
of the package neuralnet for
optional additional parameters and details about objects of class nn
.
Examples
# creating a quarry
q <- quarry(geological.context = 2, hydrogeological.context = 4,
piezometric.context = 1, quarry.position = 4,
production.catchment = 4, quality.catchment = 3)
# computing the interaction index using the default neural network
inter.idx <- compute.ann(q)
Discrete Choice Model-based Interaction Index for a Quarry
Description
Given an object of type quarry
, the function computes the probabilities
of each level of interaction (low, medium, high and very high) using a Logit
discrete choice model.
Usage
## S3 method for class 'quarry'
compute.dc(x, ...)
Arguments
x |
An object of type quarry. |
... |
Further arguments passed to or from other methods. |
Details
The model parameters have been estimated with BIOGEME and has an adjusted
\rho^2
of 0.609. The model is fully detailed in the paper "Interaction
prediction between groundwater and quarry extension using discrete choice
models and artificial neural networks" (Barthelemy et al., 2016).
Value
A list whose elements are:
p.low |
The probability of a low interaction level. |
p.medium |
The probability of a medium interaction level. |
p.high |
The probability of a high interaction level. |
p.very.high |
The probability of a very high interaciton level. |
Note
In order to use the function, the quarry must have been constructed with the
function quarry
using the default parameters.
Author(s)
Johan Barthelemy.
Maintainer: Johan Barthelemy johan@uow.edu.au.
References
Barthelemy, J., Carletti, T., Collier L., Hallet, V., Moriame, M., Sartenaer, A. (2016) Interaction prediction between groundwater and quarry extension using discrete choice models and artificial neural networks Environmental Earth Sciences (in press)
Collier, L., Barthelemy, J., Carletti, T., Moriame, M., Sartenaer, A., Hallet, V. (2015) Calculation of an Interaction Index between the Extractive Activity and Groundwater Resources Energy Procedia 76, 412-420
Bierlaire, M. (2003) BIOGEME: a free package for the estimation of discrete choice models. Swiss Transport Research Conference TRANSP-OR-CONF-2006-048
See Also
compute.ann
to compute an interaction index
based on an artificial neural network and
compute.interaction
to predict the
interaction between between the quarry and the groundwater using both the
discrete choice-based index and the neural network-based index.
Examples
# creating a quarry
q <- quarry(geological.context = 2, hydrogeological.context = 4,
piezometric.context = 1, quarry.position = 4,
production.catchment = 4, quality.catchment = 3)
# computing the interaction index
inter.idx <- compute.dc(q)
Compute the Interaction Index for a Quarry
Description
Computes the interaction indices between a quarry and groundwater and returns
an object of class interaction.index
. The function implements 2 indices
based on hazard and vulnerability parameters. The levels of each index are low,
medium, high and very high.
Usage
## S3 method for class 'quarry'
compute.interaction(x, method = "all", fun.ann = compute.ann,
fun.dc = compute.dc, ...)
Arguments
x |
A quarry object. |
method |
A string specifying the method to use for computing the index.
Possible values are: |
fun.ann |
Specify which function to be used for the neural network method.
By default it is |
fun.dc |
Specify which function to be used for the discrete choice method.
By default it is |
... |
Further arguments passed to or from other methods. For instance,
For instance the parameter |
Value
An interaction.index object consisting of a list whose elements are
method |
The selected method for computing the interaction index. |
dc |
The discrete choice-based interaction index computed by the function
specified in |
ann |
The neural network-based interaction index computed by the function
specified in |
call |
The matched call. |
Author(s)
Johan Barthelemy.
Maintainer: Johan Barthelemy johan@uow.edu.au.
References
Barthelemy, J., Carletti, T., Collier L., Hallet, V., Moriame, M., Sartenaer, A. (2016) Interaction prediction between groundwater and quarry extension using discrete choice models and artificial neural networks Environmental Earth Sciences (in press)
Collier, L., Barthelemy, J., Carletti, T., Moriame, M., Sartenaer, A., Hallet, V. (2015) Calculation of an Interaction Index between the Extractive Activity and Groundwater Resources Energy Procedia 76, 412-420
Bierlaire, M. (2003) BIOGEME: a free package for the estimation of discrete choice models. Swiss Transport Research Conference TRANSP-OR-CONF-2006-048
Krieselm, D. (2007) A Brief Introduction to Neural Networks. On-line available at http://www.dkriesel.com
Ripley, B. (1996) Pattern recognition and neural networks Cambridge university press
See Also
compute.ann
and
compute.dc
for more details about the
different interaction indices and the outputs of the corresponding functions.
train.ann
to train a neural network that
can be used with this function.
print
to print a interaction.index
object.
Examples
# creating a quarry
q <- quarry(geological.context = 2, hydrogeological.context = 4,
piezometric.context = 1, quarry.position = 4,
production.catchment = 4, quality.catchment = 3)
# computing the interaction index
inter.idx <- compute.interaction(q)
print(inter.idx)
Check if an object is an integer in a given range
Description
This function checks if an object is an integer in the range
[low.bound, up.bound]
.
Usage
int.in.range(x, low.bound = 1, up.bound = 4)
Arguments
x |
An object to be tested. |
low.bound |
The lower bound. Default value is 1. |
up.bound |
The upper bound. Default value is 4. |
Value
TRUE
if x is an integer in [low.bound, up.bound]
,
FALSE
otherwise.
Author(s)
Johan Barthelemy.
Maintainer: Johan Barthelemy johan@uow.edu.au.
Examples
# test if x in [0,10]
x <- 5
int.in.range(x, low.bound = 0, up.bound = 10)
# test if y in [1,4]
y <- 10
int.in.range(y)
Quarries and Groundwater Interaction
Description
Each record of the data frame details a quarry and a groudwater as well as their level of interaction. This level of interaction correspond to the environmental risk, or potential impact of the quarry on the regional hydrology.
Depending on the parameters of the quarry and the groundwater, there will be a low, medium, high or very high level of interaction. Hence, the interaction determines the level of investigation of the feasibility study to undertake before considering any extension of the quarry:
low hydrogeology characterization;
medium
low
+ piezometric monitoring;high
medium
+ steady state (static) mathematical model;very high
high
+ transient state (dynamic) mathematical model.
The 6 retained parameters are grouped in 2 categories:
the geological, hydrogeological and piezometric contexts defining the hazard that a quarry represents;
the relative position of the quarry and the water catchments, the production of the catchments and the potential quality of the groundwater characterising the vulnerability of the groundwater resources.
Each of those parameters is classified into 4 modalities.
Usage
data("quarries")
Format
A data frame with 3327 observations on the following 36 variables.
ID
-
The id of the observation.
INTERACTION
-
An integer in [1, 4] giving the interaction level.
Low.
Medium.
High.
Very high.
H
-
An integer in [1, 4]. Hydrogeological context. Refers to the combinations of geological formations according to their hydrodynamic characteristics.
Aquiclude formation.
Aquitard formation.
Aquifer formation.
Carbonate aquifer formation.
Z
-
An integer in [1, 4]. Piezometric context: altimetric level of the quarry floor. Characterizes the relative position between the quarry pit bottom and the groundwater piezometric level.
Higher than the piezometric level of the water table.
Lower than the piezometric level of the water table but higher than the river thalweg which is the regional base level.
Lower than the piezometric level of the water table and the altimetric level of the river thalweg which is the regional base level.
Lower than the piezometric level of the water table and the altimetric level of the river thalweg which is not the regional level any more (the river is perched).
G
-
An integer in [1, 4]. Geological context. Characterizes the lithology and extension of the geological formation exploited in the quarry and those of the neighbouring geological formations that will govern the groundwater flow directions.
Completely isolated by other formations with low permeability.
Limited extension and partly compartmentalized.
Local extension.
Regional extension.
C
-
An integer in [1, 4]. Relative position of the quarry and the water catchments. Catchments (well, spring, gallery, etc.) for public distribution of drinking water are threatened by various sources of pollution. Closer a quarry gets to the catchment, greater its impact may be important. Consequently, 4 successive perimeters, within which the activities and facilities are regulated, are set up around the catchment based on the velocity of groundwater (transfert time).
Outside the drainage zone of a catchment.
In the drainage zone of a catchment.
In the distant prevention area of a catchment (50 days of delay in case of aquifer contamination).
In the close prevention area of a catchment (24 hours of delay in case of aquifer contamination).
T
-
An integer in [1, 4]. Production of the catchments. Volume exploited in catchments for public distribution in the hydrogeological formation near the quarry.
Lower than 2 m3/h.
Between 2 and 10 m3/h.
Between 10 and 30 m3/h.
Greater than 30 m3/h.
L
-
An integer in [1, 4]. Potential quality of the catchments. Quality and the potability of the groundwater.
Poor quality.
Water potabilisable with minor treatment.
Good quality water.
Water of exceptional quality (mineral water).
low
-
An integer set to 1 if
INTERACTION
= 1, 0 otherwhise. medium
-
An integer set to 1 if
INTERACTION
= 2, 0 otherwhise. high
-
An integer set to 1 if
INTERACTION
= 3, 0 otherwhise. very.high
-
An integer set to 1 if
INTERACTION
= 4, 0 otherwhise. H1
-
An integer set to 1 if
H
= 1, 0 otherwhise. H2
-
An integer set to 1 if
H
= 2, 0 otherwhise. H3
-
An integer set to 1 if
H
= 3, 0 otherwhise. H4
-
An integer set to 1 if
H
= 4, 0 otherwhise. Z1
-
An integer set to 1 if
Z
= 1, 0 otherwhise. Z2
-
An integer set to 1 if
Z
= 2, 0 otherwhise. Z3
-
An integer set to 1 if
Z
= 3, 0 otherwhise. Z4
-
An integer set to 1 if
Z
= 4, 0 otherwhise. G1
-
An integer set to 1 if
G
= 1, 0 otherwhise. G2
-
An integer set to 1 if
G
= 2, 0 otherwhise. G3
-
An integer set to 1 if
G
= 3, 0 otherwhise. G4
-
An integer set to 1 if
G
= 4, 0 otherwhise. C1
-
An integer set to 1 if
C
= 1, 0 otherwhise. C2
-
An integer set to 1 if
C
= 2, 0 otherwhise. C3
-
An integer set to 1 if
C
= 3, 0 otherwhise. C4
-
An integer set to 1 if
C
= 4, 0 otherwhise. T1
-
An integer set to 1 if
T
= 1, 0 otherwhise. T2
-
An integer set to 1 if
T
= 2, 0 otherwhise. T3
-
An integer set to 1 if
T
= 3, 0 otherwhise. T4
-
An integer set to 1 if
T
= 4, 0 otherwhise. L1
-
An integer set to 1 if
L
= 1, 0 otherwhise. L2
-
An integer set to 1 if
L
= 2, 0 otherwhise. L3
-
An integer set to 1 if
L
= 3, 0 otherwhise. L4
-
An integer set to 1 if
L
= 4, 0 otherwhise.
Source
Deparment of Geology, University of Namur, Namur, Belgium - https://www.unamur.be/en/sci/geology
References
Barthelemy, J., Carletti, T., Collier L., Hallet,V., Moriame, M., Sartenaer, A. (2016) Interaction prediction between groundwater and quarry extension using discrete choice models and artificial neural networks Environmental Earth Sciences (in press)
Collier, L., Barthelemy, J., Carletti, T., Moriame, M., Sartenaer, A., Hallet, V. (2015) Calculation of an Interaction Index between the Extractive Activity and Groundwater Resources Energy Procedia 76, 412-420
Examples
data(quarries)
# histogram of INTERACTION levels
barplot(table(quarries$INTERACTION))
Create a Quarry Object
Description
This function create a S3 object of type quarry
characterizing a quarry
and the groundwater in the quarry's area.
A quarry
object has 6 attributes, each classified (by default) in 4
modalities coded as an integer in the range [1, 4]. These attributes are grouped
in 2 distinct categories:
the geological (
geological.context
), hydrological (hydrogeological.context
) and piezometric (piezometric.context
) contexts defining the hazard that a quarry represents;the relative position of the quarry and the water catchments (
quarry.position
), the production of the catchments (production.catchment
) and the potential quality (quality.catchment
)of the groundwater characterising the vulnerability of the groundwater ressources.
The possible values for each attribue are described in the Argument Section.
Usage
quarry(geological.context = 1, hydrogeological.context = 1,
piezometric.context = 1, quarry.position = 1, production.catchment = 1,
quality.catchment = 1, ...)
## S3 method for class 'quarry'
print(x, verbose = FALSE, ...)
Arguments
geological.context |
The geological context of the quarry. Characterizes the lithology and extension of the geological formation exploited in the quarry and those of the neighbouring geological formations that will govern the groundwater flow directions. By default the possible values are:
|
hydrogeological.context |
The hydrogeological context of the quarry. Refers to the combinations of geological formations according to their hydrodynamic characteristics. By default the possible values are:
|
piezometric.context |
The piezometric context of the quarry, i.e. the altimetric level of the quarry floor. Characterizes the relative position between the quarry pit bottom and the groundwater piezometric level. By default the possible values are:
|
quarry.position |
Relative position of the quarry and the water catchments. By default the possible values are:
|
production.catchment |
Production of the catchments. Volume exploited in catchments for public distribution in the hydrogeological formation near the quarry. By default the possible values are:
|
quality.catchment |
Potential quality of the catchments. Quality and the potability of the groundwater. By default the possible values are:
|
verbose |
If set to TRUE, then the |
... |
Further arguments passed to or from other methods. For instance if
the values of the variables must be in the range |
x |
An object of type |
Value
A quarry object consisting of a list whose elements are
G |
The geological context of the quarry. |
H |
The hydrogeological context of the quarry. |
Z |
The piezometric context of the quarry. |
C |
The relative position of the quarry and the water catchment. |
T |
The production of the water catchments. |
L |
The potential quality of the water catchments. |
G.dummy |
A vector of binary components for the dummy variable encoding of |
H.dummy |
A vector of binary components for the dummy variable encoding of |
Z.dummy |
A vector of binary components for the dummy variable encoding of |
C.dummy |
A vector of binary components for the dummy variable encoding of |
T.dummy |
A vector of binary components for the dummy variable encoding of |
L.dummy |
A vector of binary components for the dummy variable encoding of |
Note
By default, the modalities are integer in [1,4], but that can be changed
to be an integer in any given range by passing in ...
the arguments
low.bound
and up.bound
of the function
int.in.range
.
Author(s)
Johan Barthelemy.
Maintainer: Johan Barthelemy johan@uow.edu.au.
References
Barthelemy, J., Carletti, T., Collier L., Hallet, V., Moriame, M., Sartenaer, A. (2016) Interaction prediction between groundwater and quarry extension using discrete choice models and artificial neural networks Environmental Earth Sciences (sumbitted)
Collier, L., Barthelemy, J., Carletti, T., Moriame, M., Sartenaer, A., Hallet, V. (2015) Calculation of an Interaction Index between the Extractive Activity and Groundwater Resources Energy Procedia 76, 412-420
See Also
int.in.range
to a use custom range for the
values of the paramters.
print
to print a quarry
object.
as.data.frame
to coerce a quarry to
a data frame.
compute.interaction
to predict the
interaction between between the quarry and the groundwater.
Examples
# creating a quarry for which every parameter is within the default range
q1 <- quarry(geological.context = 2, hydrogeological.context = 4,
piezometric.context = 1, quarry.position = 4,
production.catchment = 4, quality.catchment = 3)
print(q1, verbose = TRUE)
# creating a quarry for which the parameters are within a custom range [0, 10]
q2 <- quarry(geological.context = 8, hydrogeological.context = 4,
piezometric.context = 0, quarry.position = 4,
production.catchment = 6, quality.catchment = 3,
low.bound = 0, up.bound = 10)
print(q2)
Training an Artificial Neural Network for Interaction Prediction.
Description
The function trains a neural network to be used with the functions
compute.interaction
and
compute.ann
. The neural network can then
be used to predict whether the level of interaction between a quarry and the
groundwater is low, medium, high or very high.
The user can specify:
the explanatory variables to be used;
the data frame used to train and validate the network;
the structure of the hidden layers;
the number of repetitions for the neural network training.
Usage
train.ann(var = c("H", "Z", "G", "C", "T", "L"), data = quarrint::quarries,
hidden = 7, rep = 1, ...)
Arguments
var |
The explanatory variable to be used. By default, all the variables in the
default data frame are used. Note that the variables must be categorical
(coded with integers) and will be transformed in dummy variables. For
instance if |
data |
The training and validation dataframe. It must contain the
variables listed in |
A vector of integer detailing the structure of the hidden layers. For instance if we want 2 hidden layers with 4 and 2 nodes respectively, then it must be it to (2, 4). The default is 7, i.e. 1 hidden layer of 7 nodes. | |
rep |
The number of repetition of the neural network to be computed. |
... |
Further arguments passed to or from other methods. See the documentation of "neuralnet" from the package "neuralnet". |
Value
A list whose elements are:
ann |
A |
prop.correct.prediction |
A list detailing for each repetition of the neural network the proportion of correct predictions. |
Author(s)
Johan Barthelemy.
Maintainer: Johan Barthelemy johan@uow.edu.au.
References
Barthelemy, J., Carletti, T., Collier L., Hallet, V., Moriame, M., Sartenaer, A. (2016) Interaction prediction between groundwater and quarry extension using discrete choice models and artificial neural networks Environmental Earth Sciences (in press)
Collier, L., Barthelemy, J., Carletti, T., Moriame, M., Sartenaer, A., Hallet, V. (2015) Calculation of an Interaction Index between the Extractive Activity and Groundwater Resources Energy Procedia 76, 412-420
Krieselm, D. (2007) A Brief Introduction to Neural Networks. On-line available at http://www.dkriesel.com
Ripley, B. (1996) Pattern recognition and neural networks Cambridge university press
See Also
The function relies on the function neuralnet
of the neuralnet package to generate an object of type nn
containing the trained neural network.
compute.interaction
and
compute.ann
to use the trained neural
network.
The data frame quarries
.
Examples
## Not run:
# training a neural network using the attribues H and T as predictors,
# with 2 hidden layers of 2 nodes each and computing 2 replications
r.ann <- train.ann(var = c("H", "T", "L", "Z"), hidden = c(2, 2), rep = 2)
# using the ann to compute the interaction
q <- quarry(geological.context = 2, hydrogeological.context = 4,
piezometric.context = 1, quarry.position = 4,
production.catchment = 4, quality.catchment = 3)
compute.interaction(q, method="ann", ann = r.ann$ann)
## End(Not run)