Custom knn-Analysis

Training Data

File Class-attribute position Datasets total 0 records x 0 attributes Parse file

Test Data

File Class-attribute position (None = -1) Datasets total 0 records x 0 attributes Parse file

General

Visualize the dataset as XY-tupels
(e.g. x₁,y₁,x₂,y₂, ... x_n, y_n) Amount of different classes 0

Examples

Character-recognition example

Training: 7,495 records x 17 attributes
Test: 3,498 records x 17 attributes
Classes: 10

Oil example

Training: 208 records x 9 attributes
Test: 208 records x 9 attributes
Classes: 9

Random data

Training: 10000 records x 40 attributes
Test: 10000 records x 40 attributes
Classes: 9

Loading example data

Parameters: k (amount of neighbours) σ (sigma) Kernel-Function None Rectangle Triangle Tricubic Gaussian Distance-Function Euclidean Manhattan Auto-increment σ if necessary to reach k neighbours

Visualization of neighbours, sigma and distances

Legend & Details

Nearest Neighbours data representation and it's weight (limited to 250 neighbours)