Skip to main content

Table 3 Mean computation times (seconds) for MLR forecasting models training and evaluation using different caching strategies

From: Concept and benchmark results for Big Data energy forecasting based on Apache Spark

Tests Training Evaluation
CS1 CS2 CS3 CS1 CS2 CS3
\(SC8_{6M}\) 356.695 (4.019) 2.822 (0.060) 2.925 (0.047) 233.737 (2.003) 5.420 (0.101) 9.467 (0.364)
\(SCl_{6M}\) 37.562 (0.290) 1.133 (0.055) 1.128 (0.041) 47.532 (0.374) 1.747 (0.050) 1.777 (0.055)
  1. The values in parenthesis are the standard deviation values
  2. Italic: lowest computation time for a given caching strategy