K-Means initialization
The K-Means initialization algorithm receives 
feature vectors as input
and chooses 
initial centroids. After initialization, K-Means algorithm
uses the initialization result to partition input data into clusters.
feature vectors as input
and chooses initial centroids. After initialization, K-Means algorithm
uses the initialization result to partition input data into clusters.Operation | Computational methods | Programming Interface |
Mathematical formulation
Computing
Given the training set 
of
-dimensional feature vectors and a positive integer , the
problem is to find a set 
of
-dimensional initial centroids.
of
-dimensional feature vectors and a positive integer , the
problem is to find a set of
-dimensional initial centroids.Computing method:
dense
The method chooses first 
feature vectors from the training set
.
feature vectors from the training set
.Programming Interface
Refer to API Reference: K-Means initialization.
Usage example
Computing
table run_compute(const table& data) {
const auto kmeans_desc = kmeans_init::descriptor<float,
kmeans_init::method::dense>{}
.set_cluster_count(10)
const auto result = compute(kmeans_desc, data);
print_table("centroids", result.get_centroids());
return result.get_centroids();
}Examples
oneAPI DPC++
Batch Processing:
oneAPI C++
Batch Processing: