Table 1 Characteristics and descriptions of datasets for regression from UCI and University of Porto repositories used for comparisons
From: Extreme learning machine versus classical feedforward network
Dataset | No of samples | No of features | Min. value of target variable | Max. value of target variable | Description |
|---|---|---|---|---|---|
Auto MPG | 392 | 21 | 9.0 | 46.6 | Estimation of car fuel consumption. Five input variables are continuous, remaining 16 features represent one-hot encodings of two discrete variables |
Bank | 8192 | 8 | 0.0 | 0.8 | Estimation of the number of customers who decide to quit the queue |
California Housing | 20,640 | 8 | 0.15 | 5.0 | Median prediction of housing prices in a neighbourhood described by 8 input variables. Dataset acquired from StatLib repository [29] |
Delta Ailerons | 7129 | 5 | \(-2.1\) \(\cdot 10^{-3}\) | 2.2 \(\cdot 10^{-3}\) | Prediction of variance of ailerons positioning in the F-16 fighter aircraft |
Machine CPU | 209 | 6 | 6 | 1150 | Prediction of multiple CPUs performance based on their specifications |
Servo | 167 | 19 | 0.13 | 7.1 | Prediction of latency in a servo system |