Chapter 18 Bosch Production Line Performance

This competition was hosted in 2016 by Bosch at https://www.kaggle.com/c/bosch-production-line-performance and was subtitled “Reduce manufacturing failures”

The price money was $30,000, the competition ended 11.11.2016 and 1373 teams had submitted a solution.

Challenge:

  • Predict internal failures
  • Data anonymized
    • measurements
    • tests
  • Benefit
    • reduce manufacturing failures

The metric was Matthews correlation coefficient (MCC) between the predicted and the observed response. The MCC is given by:

\[MCC = score=\frac{\left(TP*TN\right) - \left(FP*FN\right) }{\sqrt{\left(TP*FP\right) \left(TP*FN\right) \left(TN*FP\right) \left(TN*FN\right)}} \] where:

  • TP: number of true postive
  • FP: number of false positive
  • TN: number of true negative
  • FN: number of false negative

The best possible score for perfect predictions is 1. The leader board look as follows:

  1. 0.52401
  2. 0.51847
  3. 0.51621

The winning solution is described in the next chapter