Skip to main content

Table 2 Considered KPIs, prediction methods and metrics, regressors and sampling methods

From: Machine learning-based dynamic frequency and bandwidth allocation in self-organized LTE dense small cell deployments

KPIs

Network throughput; user throughput

Prediction methods

Bagging tree (TBAG, Treebag)

 

Boosted tree (BTR)

 

Kohonen network (KOH)

 

SVM radial (SVM)

 

K-nearest neighbor (KNN)

 

Projection pursuit regression (PPR)

 

Linear (LM)

Prediction metrics

RMSE user fit; 95th percentile RMSE

Regressors

SINR; SINR/MAC throughput mapping

Sampling

Random; stratified