Teradata R Package Function Reference | 17.00 - 17.00 - SVMSparse - Teradata R Package

Teradata® R Package Function Reference

prodname
Teradata R Package
vrm_release
17.00
created_date
September 2020
category
Programming Reference
featnum
B700-4007-090K

Description

The SVMSparse function takes training data (in sparse format) and outputs a predictive model in binary format, which is input to the functions SVMSparsePredict td_svm_sparse_predict_mle and SVMSparseSummary td_svm_sparse_summary_mle.

Usage

  td_svm_sparse_mle (
      data = NULL,
      sample.id.column = NULL,
      attribute.column = NULL,
      value.column = NULL,
      label.column = NULL,
      cost = 1.0,
      bias = 0.0,
      hash = FALSE,
      hash.buckets = NULL,
      class.weights = NULL,
      max.step = 100,
      epsilon = 0.01,
      seed = 0,
      data.sequence.column = NULL
  )

Arguments

data

Required Argument.
Specifies the name of the tbl_teradata that contains the training samples.

sample.id.column

Required Argument.
Specifies the name of the input column that contains the identifiers of the training samples.
Types: character

attribute.column

Required Argument.
Specifies the name of the input column that contains the attributes of the samples.
Types: character

value.column

Optional Argument.
Specifies the name of the input column that contains the attribute values. By default, each attribute has the value 1.
Types: character

label.column

Required Argument.
Specifies the name of the input column that contains the classes of the samples.
Types: character

cost

Optional Argument.
Speifies the regularization parameter in the SVM soft-margin loss function. The cost must be greater than 0.0.
Default Value: 1.0
Types: numeric

bias

Optional Argument.
Specifies a non-negative value. If the value is greater than zero, each sample x in the training set will be converted to (x, b); that is, it will add another dimension containing the bias value b. This argument addresses situations where not all samples center at 0.
Default Value: 0.0
Types: numeric

hash

Optional Argument.
Specifies whether to use hash projection on attributes. Hash projection can accelerate processing speed but can slightly decrease accuracy.
Note: You must use hash projection if the dataset has more features than fit into memory.
Default Value: FALSE
Types: logical

hash.buckets

Optional Argument.
Valid only if hash is TRUE. Specifies the number of buckets for hash projection. In most cases, the function can determine the appropriate number of buckets from the scale of the input data set. However, if the dataset has a very large number of features, you might have to specify number of buckets to accelerate the function.
Types: integer

class.weights

Optional Argument.
Specifies the weights for different classes. The format is: "classlabel m:weight m".
For a single class, the weight can be specified as the value of type character:
"classlabel m:weight m"
For multiple classes, the weights can be specified as a vector of characters:
c("classlabel m:weight m", "classlabel n:weight n")
If weight for a class is given, the cost parameter for this class is weight * cost. A weight larger than 1 often increases the accuracy of the corresponding class; however, it may decrease global accuracy. Classes not assigned a weight in this argument are assigned a weight of 1.0.
Types: character OR vector of characters

max.step

Optional Argument.
A positive integer value that specifies the maximum number of iterations of the training process. One step means that each sample is seen once by the trainer. The input value must be in the range (0, 10000].
Default Value: 100
Types: integer

epsilon

Optional Argument.
Specifies the termination criterion.
When the difference between the values of the loss function in two sequential iterations is less than this number, the function stops.
epsilon must be greater than 0.0.
Default Value: 0.01
Types: numeric

seed

Optional Argument.
Specifies a long integer value used to order the training set randomly and consistently. This value can be used to ensure that the same model will be generated if the function is run multiple times in a given database with the same arguments. The input value must be in the range [0, 9223372036854775807].
Default Value: 0
Types: numeric

data.sequence.column

Optional Argument.
Specifies the vector of column(s) that uniquely identifies each row of the input argument "data". The argument is used to ensure deterministic results for functions which produce results that vary from run to run.
Types: character OR vector of Strings (character)

Value

Function returns an object of class "td_svm_sparse_mle" which is a named list containing objects of class "tbl_teradata".
Named list members can be referenced directly with the "$" operator using following names:

  1. model.table

  2. output

Examples

    # Get the current context/connection
    con <- td_get_context()$connection
    
    # Load example data.
    loadExampleData("svmsparse_example", "svm_iris_input_train")

    # Create object(s) of class "tbl_teradata".
    svm_iris_input_train <- tbl(con, "svm_iris_input_train")

    # Example -
    td_svm_sparse_out <- td_svm_sparse_mle(data = svm_iris_input_train,
                                           sample.id.column = "id",
                                           attribute.column = "attribute",
                                           value.column = "value1",
                                           label.column = "species",
                                           max.step = 150,
                                           seed = 0
                                          )