TD_​KNN

This function classifies data objects based on proximity to other data objects with known classification.

Options

Accumulate

Specify the names of the input table columns that need to be copied from the input test table to output.

EmitDistances

A boolean flag which specifies whether the neighbor distances are to be emitted in the output.

EmitNeighbors

A boolean flag which specifies whether the neighbors are to be emitted in the output.

IDColumn

Specify the name of the column that uniquely identifies a data object both in training table and test table.

InputColumns

Specify the names of the training table columns that the function uses to compute the distance between a test object and the training objects. The test table must also have these columns.

K

Specify the number of nearest neighbors to use in the algorithm. Any positive integer value > 0 and <= 1000 can be chosen.

ModelType

Specify the Model Type for KNN function. Acceptable values are 'regression', 'classification', or 'neighbors'.

OutputProb

Indicates whether the function should output the probability for each response specified in 'Responses'. If 'Responses' is not given, outputs the probability of the predicted response.

ResponseColumn

Specify the name of the training table column that contains the numeric response variable values to be used for prediction in KNN based regression or classification.

Responses (strings seperated by new line)

Specify the class labels for which to output probabilities.

Output Schema

Output Schema, if Volatile is true then use user login as the schema.

Output Table

Output Table

VAL Location

VAL Location

Volatile

Specifies whether the table should be a VOLATILE table. If true, then the table is automatically deleted, otherwise it is users responsibility to remove or clean it up for space.

Tolerance

When a non-zero voting weight is used, the case of zero distance will cause the weight (w=1/POWER(distance, voting_weight)) to be undefined. Tolerance allows the user to define the smallest distance to be considered. For any distance under the given tolerance, the weight will be calculated as w=1/POWER(tolerance, voting_weight).

VotingWeight

Specifies the voting weight of the training object for determining the class of the test object as a function of the distance between the training and test objects. The voting weight is calculated as w, where w=1/POWER(distance, voting_weight) and distance is the distance between the test object and the training object. Must be a non-negative real number.

Nodes

• Boolean Configuration4 ×
• Column Filter Configuration4 ×
• String Configuration4 ×
• Table Row to Variable4 ×
• Double Configuration2 ×
• Show all 16 nodes

Extensions

• FeatureKNIME Base nodes
• FeatureKNIME Database
• FeatureKNIME Quick Forms

Links

• No links available