Table 1 ANOVA Table showing the number of nodes in a cluster is a statistically significant factor regarding training time training time on the 1:1 dataset size
From: A parallel and distributed stochastic gradient descent implementation using commodity clusters
| Â | Df | Sum Sq | Mean Sq | F value | Pr (>Â F) |
|---|---|---|---|---|---|
Nodes | 4 | 35,419.46 | 8854.87 | 3968.24 | 0.0000 |
Residuals | 95 | 211.99 | 2.23 | Â | Â |