fast cross validation for incremental learning
play

Fast Cross-Validation for Incremental Learning Pooria Joulani, Andr - PowerPoint PPT Presentation

Fast Cross-Validation for Incremental Learning Pooria Joulani, Andr as Gy orgy, Csaba Szepesv ari Department of Computing Science University of Alberta Edmonton, Alberta July 11, 2015 Appearing in the International Joint Conference on


  1. Fast Cross-Validation for Incremental Learning Pooria Joulani, Andr´ as Gy¨ orgy, Csaba Szepesv´ ari Department of Computing Science University of Alberta Edmonton, Alberta July 11, 2015 Appearing in the International Joint Conference on Artificial Intelligence , Buenos Aires, Argentina, July 2015.

  2. TreeCV : Fast CV for Incremental Learning A new cross-validation algorithm: TreeCV ! Joulani, Gy¨ orgy, Szepesv´ ari Fast Cross-Validation for Incremental Learning July 11, 2015 1 / 3

  3. TreeCV : Fast CV for Incremental Learning A new cross-validation algorithm: TreeCV ! Speed up CV for incremental, single-pass algorithms. Joulani, Gy¨ orgy, Szepesv´ ari Fast Cross-Validation for Incremental Learning July 11, 2015 1 / 3

  4. TreeCV : Fast CV for Incremental Learning A new cross-validation algorithm: TreeCV ! Speed up CV for incremental, single-pass algorithms. ◮ k -fold CV: running time penalty O (log k ) instead of O ( k )! Joulani, Gy¨ orgy, Szepesv´ ari Fast Cross-Validation for Incremental Learning July 11, 2015 1 / 3

  5. TreeCV : Fast CV for Incremental Learning A new cross-validation algorithm: TreeCV ! Speed up CV for incremental, single-pass algorithms. ◮ k -fold CV: running time penalty O (log k ) instead of O ( k )! ◮ Leave-One-Out in O (log n )! Joulani, Gy¨ orgy, Szepesv´ ari Fast Cross-Validation for Incremental Learning July 11, 2015 1 / 3

  6. TreeCV : Fast CV for Incremental Learning A new cross-validation algorithm: TreeCV ! Speed up CV for incremental, single-pass algorithms. ◮ k -fold CV: running time penalty O (log k ) instead of O ( k )! ◮ Leave-One-Out in O (log n )! Does not rely on a specific Joulani, Gy¨ orgy, Szepesv´ ari Fast Cross-Validation for Incremental Learning July 11, 2015 1 / 3

  7. TreeCV : Fast CV for Incremental Learning A new cross-validation algorithm: TreeCV ! Speed up CV for incremental, single-pass algorithms. ◮ k -fold CV: running time penalty O (log k ) instead of O ( k )! ◮ Leave-One-Out in O (log n )! Does not rely on a specific ◮ type of the learning problem (classification, regression, density estimation, etc.); Joulani, Gy¨ orgy, Szepesv´ ari Fast Cross-Validation for Incremental Learning July 11, 2015 1 / 3

  8. TreeCV : Fast CV for Incremental Learning A new cross-validation algorithm: TreeCV ! Speed up CV for incremental, single-pass algorithms. ◮ k -fold CV: running time penalty O (log k ) instead of O ( k )! ◮ Leave-One-Out in O (log n )! Does not rely on a specific ◮ type of the learning problem (classification, regression, density estimation, etc.); ◮ inner structure of the algorithm (e.g., QP, influence matrix, etc.); Joulani, Gy¨ orgy, Szepesv´ ari Fast Cross-Validation for Incremental Learning July 11, 2015 1 / 3

  9. TreeCV : Fast CV for Incremental Learning A new cross-validation algorithm: TreeCV ! Speed up CV for incremental, single-pass algorithms. ◮ k -fold CV: running time penalty O (log k ) instead of O ( k )! ◮ Leave-One-Out in O (log n )! Does not rely on a specific ◮ type of the learning problem (classification, regression, density estimation, etc.); ◮ inner structure of the algorithm (e.g., QP, influence matrix, etc.); ◮ loss function used for CV (accuracy, F-measure, etc.). Joulani, Gy¨ orgy, Szepesv´ ari Fast Cross-Validation for Incremental Learning July 11, 2015 1 / 3

  10. TreeCV : Fast CV for Incremental Learning A new cross-validation algorithm: TreeCV ! Speed up CV for incremental, single-pass algorithms. ◮ k -fold CV: running time penalty O (log k ) instead of O ( k )! ◮ Leave-One-Out in O (log n )! Does not rely on a specific ◮ type of the learning problem (classification, regression, density estimation, etc.); ◮ inner structure of the algorithm (e.g., QP, influence matrix, etc.); ◮ loss function used for CV (accuracy, F-measure, etc.). Easy parallelization / distributed computing. Joulani, Gy¨ orgy, Szepesv´ ari Fast Cross-Validation for Incremental Learning July 11, 2015 1 / 3

  11. TreeCV : Fast CV for Incremental Learning A new cross-validation algorithm: TreeCV ! Speed up CV for incremental, single-pass algorithms. ◮ k -fold CV: running time penalty O (log k ) instead of O ( k )! ◮ Leave-One-Out in O (log n )! Does not rely on a specific ◮ type of the learning problem (classification, regression, density estimation, etc.); ◮ inner structure of the algorithm (e.g., QP, influence matrix, etc.); ◮ loss function used for CV (accuracy, F-measure, etc.). Easy parallelization / distributed computing. Theoretical bounds and experimental results on the speed and accuracy. Joulani, Gy¨ orgy, Szepesv´ ari Fast Cross-Validation for Incremental Learning July 11, 2015 1 / 3

  12. TreeCV in action: Leave-One-Out CV estimation SVM Classification with PEGASOS (Shalev-Shwartz et al., 2011). ◮ CV over the 0-1 loss. Least-square regression with SGD (Nemirovski et al., 2009). ◮ CV over the squared loss. Joulani, Gy¨ orgy, Szepesv´ ari Fast Cross-Validation for Incremental Learning July 11, 2015 2 / 3

Recommend


More recommend