Feature Subset Selection for Learning Huge Configuration Spaces: The case of Linux Kernel Size
0 likes57 views
This document discusses feature subset selection for predicting properties of Linux kernel configurations, highlighting that out of over 9,000 options, only about 300 are essential for effective prediction. It presents findings that traditional state-of-the-art methods struggle with high-dimensional spaces, while tree-based algorithms yield low error rates and improved training times when using a reduced feature set. Furthermore, the results align with domain knowledge, revealing that many influential kernel options are poorly documented.
Feature Subset Selection for Learning Huge Configuration Spaces: The case of Linux Kernel Size
- 1. Feature Subset Selection for Learning Huge Configuration Spaces The case of Linux Kernel Size Mathieu Acher, Hugo Martin, Juliana Alves Pereira, Luc Lesoil, Arnaud Blouin, Jean-Marc Jézéquel, Djamel Eddine Khelladi, Olivier Barais Preprint: https://hal.inria.fr/hal-03720273
- 2. 15,000+ options Linux 5.2.8, arm (% of types’ options) 39000 26000 ≈106000 variants (without constraints) 2
- 4. Dimensionality reduction with feature selection Huge configuration space ≈106000 configurations Large option/feature* set: 9K+ options for x86_64 Hypothesis: only a subset of options matter when predicting properties of variants 4 *options (~Linux features) are encoded as features (~predictive variables in learning problems)
- 5. Dimensionality reduction with feature selection Huge configuration space ≈106000 configurations Large option/feature set: 9K+ options for x86_64 Hypothesis: only a subset of options matter when predicting properties of variants. Very few studies at this scale p options p’ options with p’ << p n configurations 5
- 6. Hypothesis: Only a subset of options matter when predicting properties of variants. Key results: ● Some state-of-the-art solutions are not scaling due to “too many feature interactions” (think about combinatorial with thousands of features!) ● Only ~300 features* (instead of 9K+) are sufficient to efficiently predict and even outperforms the accuracy of “learning over all features/options” ● Training time can be decreased ● Identification of influential options is consistent with, and can even improve, the expert knowledge about Linux kernel configuration. 6 *options (~Linux features) are encoded as features (~predictive variables in learning problems)
- 7. Configurable software system Configurations Variants Quantitative property (eg related to performance, security, energy consumption) 176.8Mb Linux kernel .config (compile-time/Kconfig) Kernel variants (binaries) binary size7
- 8. Configurable software system Configurations Variants Quantitative property (eg related to performance, security, energy consumption) 16.1Mb Linux kernel .config (compile-time/Kconfig) Kernel variants (binaries) binary size8
- 9. Configurable software system Configurations Variants Quantitative property (eg related to performance, security, energy consumption) 176.8Mb Linux kernel .config (compile-time/Kconfig) Kernel variants (binaries) binary size 16.1Mb 77.2Mb 9
- 10. Configurable software system Configurations Variants Quantitative property (eg related to performance, security, energy consumption) Linux kernel .config (compile-time/Kconfig) Kernel variants (binaries) binary size ? 10
- 11. Challenge: you cannot build ≈106000 configurations; sampling and learning to the rescue but… Is it accurate? Is it effective with p’ features and feature selection? How many features*? Which options* matter? 7.1Mb 176.8Mb ? 11 p’ options with p’ << p *options (~Linux features) are encoded as features (~predictive variables in learning problems)
- 12. A challenging case ● Targeted non-functional, quantitative property: binary size ○ interest for maintainers/users of the Linux kernel (embedded systems, cloud, etc.) ○ challenging to predict (cross-cutting options, interplay with compilers/build systems, etc.) ● Dataset: version 4.13.3, x86_64 arch, measurements of 95K+ random configurations ○ paranoiac about deep variability since 2017: Docker to control the build environment and scale ○ build: 8 minutes on average ○ diversity: from 7Mb to 1.9Gb 12
- 13. TUXML: Sampling, Measuring, Learning 13 Most of the work consider a relatively low number of options (<50) Linux has 9K+ options for x86_64 Feature subset selection vs recursive feature elimination: scale? accuracy? *EX: execution, SI: simulation, SA: static analysis, UF: user feedback, SM: synthetic measurements.
- 14. TUXML: Sampling, Measuring, Learning Docker for a reproducible environment with tools/packages needed and Python procedures inside Easy to launch campaign: ”python kernel_generator.py 10” builds/measures 10 random configurations (information sent to a database) https://github.com/TuxML/ 14
- 15. TUXML: Sampling, Measuring, Learning Docker for a reproducible environment with tools/packages needed and Python procedures inside Easy to launch campaign: ”python3 kernel_generator.py 10” builds/measures 10 random configurations (information sent to a database) https://github.com/TuxML/ 15
- 16. Data: version 4.13.3 (x86_64) 95K+ configurations for Linux 4.13.3 (and 15K hours of computation on a grid computing) 16
- 17. RQ1: How do SOTA techniques perform on huge configuration spaces? ● Linear-based algorithms : high error rate (it’s not additive!) ● Polynomial regression & performance-influence model : Out Of Memory (too much interactions and not designed for 9K+ options) ● Tree-based algorithms & neural networks: low error rate Mean Absolute Percentage Error (MAPE): the lower the better 17 N : percentage of the dataset used to training
- 18. Dimensionality reduction with feature selection Huge configuration space ≈106000 configurations Large options/feature set: 9K+ options for x86_64 Only a subset of options matter when predicting properties of variants. RQ2: How accurate is the prediction model with and without feature selection? p options p’ options with p’ << p n configurations 18
- 19. Dimensionality reduction with Tree-based feature selection Tree-based algorithm (Random Forest) p=8.743 options Learn on Full dataset p’ <<<<< p options Reduced dataset Filter Any learning algorithm Learn on DEBUG_INFO (0.33) active_options (0.19) group_129 (0.14) DEBUG_INFO_REDUCED (0.11) DEBUG_INFO_SPLIT (0.08) feature ranking list (based on feature importance) 19
- 20. RQ2: Tree-based Feature Selection pays off! ● Tree-based algorithms & neural networks: ○ Lower error rate ○ Lower training time ■ Random forest : 18x ■ Gradient Boosting Tree : 5x ● Simpler models, easier to train, and improved accuracy ● Bonus: interpretable and consistent with domain knowledge 20
- 21. RQ2: Optimal number of features/options when performing feature selection ● Depending on algorithm ○ Gradient Boosting Trees & Neural networks : 1500 ● Depending on training set size ● Random forest : 250 options Sweet spot where only ~300 features are sufficient to efficiently train a Random Forest and a Gradient Boosting Tree to obtain a prediction model that outperforms other baselines operating over the full set of features (6% prediction errors for 40K configurations). 21
- 22. RQ3+4: Stability of influential options and Training time reduction Using an ensemble of Random Forest allows the creation of a far more stable list, with more than 95% common features in top 300 between multiple list Tree-based feature selection speeds the model training at least 5 times up to 48 times (since p’ <<<< p) 22
- 23. RQ5: How do feature ranking lists, as computed by tree-based feature selection, relate to Linux knowledge? Top influential options 147 documented options in Kconfig 0 - 50 7 50 - 250 6 250 - 500 6 500 - 1500 28 1500 - 69 Top 50 options in the feature ranking list represents 95% of the feature importance; collinearity and interpretability: beware! Incompleteness of Linux documentation: ● Vast majority of influential options is either not documented or not referring to size: only 7 options of the top 50 are documented as having a clear influence on size ● Leveraging all the 147 options in the Linux documentation (and only them) leads to prediction error of 23.6% (instead of <6% for our feature ranking list) Relevance: Investigations and exchanges with domain experts confirm the relevance of the top 50, giving 6 categories of options. Effective identification of important features: ● consistent with Linux knowledge (Kconfig documentation and expert insight) ● can be used to refine or augment the incomplete documentation of the Linux kernel. 23
- 24. Kaggle competition using our dataset https://www.kaggle.com/competitions/linux-kernel-size/overview 24 We can benefit from contributions of the machine learning community… And our dataset/problems are raising interests.
- 25. Conclusion Feature subset selection is effective over the huge configuration space of Linux: ● only ~300 features out of 9K+ ● accuracy is better with than without tree-based ● training time is decreased ● interpretability: identification of influential options is consistent with, and can even improve, the expert knowledge about Linux kernel configuration Future work ● Replication on different versions of Linux ● Does feature ranking list transfer to other versions? https://www.kaggle.com/competitions/linux-kernel-size/overview 25
- 26. 26
- 28. Decision Tree ● Ability to handle interactions between features ● Low impact of combinatorial explosion ● Competitive accuracy ● Interpretability ○ Decision rules ○ Feature importance ● Ensembles : Random Forests, Gradient Boosting Trees... ○ More accurate, less interpretable 28
- 29. Kpredict Python module for Python 3.8+ ( https://github.com/HugoJPMartin/kpredict ) Works for many kernel versions and any configuration x86_64 Error : ≃ 6.3% 97% of the predictions are below 20% error H. Martin, M. Acher, J. A. Pereira, L. Lesoil, J. Jézéquel and D. E. Khelladi, “Transfer learning across variants and versions: The case of linux kernel size” Transactions on Software Engineering (TSE), 2021 29
- 30. Published at IEEE Transactions on Software Engineering (TSE) in 2021 Preprint: https://hal.inria.fr/hal-03358817 30
- 31. Linux Kernel 31
- 32. Backup / Draft slides 32
- 33. Transfer learning “Inductive transfer refers to any algorithmic process by which structure or knowledge derived from a learning problem is used to enhance learning on a related problem.” - Jeremy West in A theoretical foundation for inductive transfer ● 100.000 configuration measurements, 15.000 hours of computation ● Mission Impossible : Saving Private Model 4.13 ○ Budget : 5.000 configurations measurements (one night worth of ISTIC computing power) 33
- 35. Model 4.13: Genesis f1 f2 f3 ... fn 1 0 0 ... 1 0 1 0 ... 0 ... ... ... ... ... 1 1 1 ... 0 Size 16MB 52MB ... 115MB 35
- 36. Model 4.13: Genesis f1 f2 f3 ... fn 1 0 0 ... 1 0 1 0 ... 0 ... ... ... ... ... 1 1 1 ... 0 Size 16MB 52MB ... 115MB 36
- 37. Model 4.13: Genesis f1 f2 f3 ... fn 1 0 0 ... 1 0 1 0 ... 0 ... ... ... ... ... 1 1 1 ... 0 Size 16MB 52MB ... 115MB Gradient Boosting Tree algorithm Features Target 37
- 38. Model 4.13: Genesis f1 f2 f3 ... fn 1 0 0 ... 1 0 1 0 ... 0 ... ... ... ... ... 1 1 1 ... 0 Size 16MB 52MB ... 115MB Gradient Boosting Tree algorithm Features Target Model 4.13 38
- 39. Model 4.13: Genesis f1 f2 f3 ... fn 1 0 0 ... 1 0 1 0 ... 0 ... ... ... ... ... 1 1 1 ... 0 Size 16MB 52MB ... 115MB Gradient Boosting Tree algorithm Features Target Model 4.13 f1 f2 f3 ... fn 0 1 1 ... 0 1 0 0 ... 1 ... ... ... ... ... 1 0 1 ... 0 39
- 40. Model 4.13: Genesis f1 f2 f3 ... fn 1 0 0 ... 1 0 1 0 ... 0 ... ... ... ... ... 1 1 1 ... 0 Size 16MB 52MB ... 115MB Gradient Boosting Tree algorithm Features Target Model 4.13 Size 18MB 25MB ... 228MB Predict f1 f2 f3 ... fn 0 1 1 ... 0 1 0 0 ... 1 ... ... ... ... ... 1 0 1 ... 0 40
- 41. Model 4.13: Genesis f1 f2 f3 ... fn 1 0 0 ... 1 0 1 0 ... 0 ... ... ... ... ... 1 1 1 ... 0 Size 16MB 52MB ... 115MB Gradient Boosting Tree algorithm Features Target Model 4.13 Size 18MB 25MB ... 228MB Predict ✅ ✅ ✅ f1 f2 f3 ... fn 0 1 1 ... 0 1 0 0 ... 1 ... ... ... ... ... 1 0 1 ... 0 41
- 43. Model Shifting f1 f2 f3 ... fn 1 0 0 ... 1 0 1 0 ... 0 ... ... ... ... ... 1 1 1 ... 0 Size 22MB 68MB ... 105MB Gradient Boosting Tree algorithm Features Target Model 4.15 43
- 44. Model Shifting f1 f2 f3 ... fn 1 0 0 ... 1 0 1 0 ... 0 ... ... ... ... ... 1 1 1 ... 0 Size 22MB 68MB ... 105MB Gradient Boosting Tree algorithm Features Target Model 4.15 Size 19MB 26MB ... 298MB Predict ❌ ❌ ✅ f1 f2 f3 ... fn 0 1 1 ... 0 1 0 0 ... 1 ... ... ... ... ... 1 0 1 ... 0 44
- 45. Model Shifting f1 f2 f3 ... fn 1 0 0 ... 1 0 1 0 ... 0 ... ... ... ... ... 1 1 1 ... 0 Size 22MB 68MB ... 105MB Gradient Boosting Tree algorithm Features Target Model 4.15 Size 19MB 26MB ... 298MB Predict ❌ ❌ ✅ f1 f2 f3 ... fn 0 1 1 ... 0 1 0 0 ... 1 ... ... ... ... ... 1 0 1 ... 0 Model 4.13 45
- 46. Model Shifting f1 f2 f3 ... fn 1 0 0 ... 1 0 1 0 ... 0 ... ... ... ... ... 1 1 1 ... 0 Size 22MB 68MB ... 105MB Gradient Boosting Tree algorithm Features Target Model 4.15 f1 f2 f3 ... fn 0 1 1 ... 0 1 0 0 ... 1 ... ... ... ... ... 1 0 1 ... 0 Size 19MB 26MB ... 298MB Predict ❌ ❌ ✅ Model 4.13 Old Size 16MB 52MB ... 115MB Old Size 18MB 25MB ... 228MB Predict Predict 46
- 47. Model Shifting f1 f2 f3 ... fn 1 0 0 ... 1 0 1 0 ... 0 ... ... ... ... ... 1 1 1 ... 0 Size 22MB 68MB ... 105MB Gradient Boosting Tree algorithm Features Target Shifting Model 4.15 f1 f2 f3 ... fn 0 1 1 ... 0 1 0 0 ... 1 ... ... ... ... ... 1 0 1 ... 0 Size 21MB 35MB ... 298MB Predict ✅ ✅ ✅ Model 4.13 Old Size 16MB 52MB ... 115MB Old Size 18MB 25MB ... 228MB Predict Predict 47
- 48. Results 48
- 49. Results Budget : 5.000 configurations ● Model shifting : ○ From 5.6% to 7.1% error rate ● Scratch : ○ From 8.2% to 9.2% error rate 49
- 50. Results Budget : 5.000 configurations ● Model shifting : ○ From 5.6% to 7.1% error rate ● Scratch : ○ From 8.2% to 9.2% error rate Budget : 10.000 configurations ● Model shifting : ○ From 5.2% to 6.1% error rate ● Scratch : ○ From 7.1% to 7.7% error rate 50
- 51. Results Budget : 1.000 configurations ● Model shifting : ○ From 6.7% to 10.6% error rate ● Scratch : ○ From 14.9% to 16.7% error rate Budget : 5.000 configurations ● Model shifting : ○ From 5.6% to 7.1% error rate ● Scratch : ○ From 8.2% to 9.2% error rate Budget : 10.000 configurations ● Model shifting : ○ From 5.2% to 6.1% error rate ● Scratch : ○ From 7.1% to 7.7% error rate 51
- 53. Incremental Model Shifting Model 4.13 + Shifting Model 4.15 = Model 4.15 Model 4.13 + Shifting Model 4.20 = Model 4.20 Model 4.13 + Shifting Model 5.0 = Model 5.0 Model 4.13 + Shifting Model 5.4 = Model 5.4 Model 4.13 + Shifting Model 5.7 = Model 5.7 Model 4.13 + Shifting Model 5.8 = Model 5.8 Source + Shifting Model = Full Model Simple Model Shifting 53
- 54. Incremental Model Shifting Model 4.13 + Shifting Model 4.15 = Model 4.15 Model 4.13 + Shifting Model 4.20 = Model 4.20 Model 4.13 + Shifting Model 5.0 = Model 5.0 Model 4.13 + Shifting Model 5.4 = Model 5.4 Model 4.13 + Shifting Model 5.7 = Model 5.7 Model 4.13 + Shifting Model 5.8 = Model 5.8 Source + Shifting Model = Full Model Simple Model Shifting Model 4.13 + Shifting Model 4.15 = Model 4.15 Model 4.13 + Shifting Model 4.20 = Model 4.20 Source + Shifting Model = Full Model Incremental Model Shifting 54
- 55. Incremental Model Shifting Model 4.13 + Shifting Model 4.15 = Model 4.15 Model 4.13 + Shifting Model 4.20 = Model 4.20 Model 4.13 + Shifting Model 5.0 = Model 5.0 Model 4.13 + Shifting Model 5.4 = Model 5.4 Model 4.13 + Shifting Model 5.7 = Model 5.7 Model 4.13 + Shifting Model 5.8 = Model 5.8 Source + Shifting Model = Full Model Simple Model Shifting Model 4.13 + Shifting Model 4.15 = Model 4.15 Model 4.15 + Shifting Model 4.20 = Model 4.20 Source + Shifting Model = Full Model 9 Incremental Model Shifting 55
- 56. Incremental Model Shifting Model 4.13 + Shifting Model 4.15 = Model 4.15 Model 4.13 + Shifting Model 4.20 = Model 4.20 Model 4.13 + Shifting Model 5.0 = Model 5.0 Model 4.13 + Shifting Model 5.4 = Model 5.4 Model 4.13 + Shifting Model 5.7 = Model 5.7 Model 4.13 + Shifting Model 5.8 = Model 5.8 Source + Shifting Model = Full Model Simple Model Shifting Model 4.13 + Shifting Model 4.15 = Model 4.15 Model 4.15 + Shifting Model 4.20 = Model 4.20 Model 4.20 + Shifting Model 5.0 = Model 5.0 Model 5.0 + Shifting Model 5.4 = Model 5.4 Model 5.4 + Shifting Model 5.7 = Model 5.7 Model 5.7 + Shifting Model 5.8 = Model 5.8 Source + Shifting Model = Full Model 9 Incremental Model Shifting 56
- 57. 10 57
- 58. Results Budget : 5.000 configurations ● Model shifting : ○ From 5.6% to 7.1% error rate ● Scratch : ○ From 8.2% to 9.2% error rate ● Incremental Shifting : ○ From 5.6% to 7.5% 10 58
- 59. Results Budget : 5.000 configurations ● Model shifting : ○ From 5.6% to 7.1% error rate ● Scratch : ○ From 8.2% to 9.2% error rate ● Incremental Shifting : ○ From 5.6% to 7.5% Budget : 10.000 configurations ● Model shifting : ○ From 5.2% to 6.1% error rate ● Scratch : ○ From 7.1% to 7.7% error rate ● Incremental Shifting : ○ From 5.2% to 6.5% 10 59
- 60. Results Budget : 1.000 configurations ● Model shifting : ○ From 6.7% to 10.6% error rate ● Scratch : ○ From 14.9% to 16.7% error rate ● Incremental Shifting : ○ From 6.7% to 13.3% Budget : 5.000 configurations ● Model shifting : ○ From 5.6% to 7.1% error rate ● Scratch : ○ From 8.2% to 9.2% error rate ● Incremental Shifting : ○ From 5.6% to 7.5% Budget : 10.000 configurations ● Model shifting : ○ From 5.2% to 6.1% error rate ● Scratch : ○ From 7.1% to 7.7% error rate ● Incremental Shifting : ○ From 5.2% to 6.5% 10 60
- 61. Results Budget : 1.000 configurations ● Model shifting : ○ From 6.7% to 10.6% error rate ● Scratch : ○ From 14.9% to 16.7% error rate ● Incremental Shifting : ○ From 6.7% to 13.3% Budget : 5.000 configurations ● Model shifting : ○ From 5.6% to 7.1% error rate ● Scratch : ○ From 8.2% to 9.2% error rate ● Incremental Shifting : ○ From 5.6% to 7.5% Budget : 10.000 configurations ● Model shifting : ○ From 5.2% to 6.1% error rate ● Scratch : ○ From 7.1% to 7.7% error rate ● Incremental Shifting : ○ From 5.2% to 6.5% Model 4.13 Budget : 85.000 configurations 10 61
- 62. Results Budget : 1.000 configurations ● Model shifting : ○ From 6.7% to 10.6% error rate ● Scratch : ○ From 14.9% to 16.7% error rate ● Incremental Shifting : ○ From 6.7% to 13.3% Budget : 5.000 configurations ● Model shifting : ○ From 5.6% to 7.1% error rate ● Scratch : ○ From 8.2% to 9.2% error rate ● Incremental Shifting : ○ From 5.6% to 7.5% Budget : 10.000 configurations ● Model shifting : ○ From 5.2% to 6.1% error rate ● Scratch : ○ From 7.1% to 7.7% error rate ● Incremental Shifting : ○ From 5.2% to 6.5% Model 4.13 Budget : 85.000 configurations Model 4.13 Budget : 20.000 configurations 10 62
- 63. Results Budget : 1.000 configurations ● Model shifting : ○ From 6.7% to 10.6% error rate ● Scratch : ○ From 14.9% to 16.7% error rate ● Incremental Shifting : ○ From 6.7% to 13.3% Budget : 5.000 configurations ● Model shifting : ○ From 5.6% to 7.1% error rate ● Scratch : ○ From 8.2% to 9.2% error rate ● Incremental Shifting : ○ From 5.6% to 7.5% Budget : 10.000 configurations ● Model shifting : ○ From 5.2% to 6.1% error rate ● Scratch : ○ From 7.1% to 7.7% error rate ● Incremental Shifting : ○ From 5.2% to 6.5% Model 4.13 Budget : 85.000 configurations Model 4.13 Budget : 20.000 configurations Budget : 5.000 configurations ● Model shifting : ○ From 6.7% to 7.9% error rate ● Scratch : ○ From 8.2% to 9.2% error rate ● Incremental Shifting : ○ From 6.7% to 7.9% 10 63
- 64. Results Budget : 1.000 configurations ● Model shifting : ○ From 6.7% to 10.6% error rate ● Scratch : ○ From 14.9% to 16.7% error rate ● Incremental Shifting : ○ From 6.7% to 13.3% Budget : 5.000 configurations ● Model shifting : ○ From 5.6% to 7.1% error rate ● Scratch : ○ From 8.2% to 9.2% error rate ● Incremental Shifting : ○ From 5.6% to 7.5% Budget : 10.000 configurations ● Model shifting : ○ From 5.2% to 6.1% error rate ● Scratch : ○ From 7.1% to 7.7% error rate ● Incremental Shifting : ○ From 5.2% to 6.5% Model 4.13 Budget : 85.000 configurations Model 4.13 Budget : 20.000 configurations Budget : 5.000 configurations ● Model shifting : ○ From 6.7% to 7.9% error rate ● Scratch : ○ From 8.2% to 9.2% error rate ● Incremental Shifting : ○ From 6.7% to 7.9% Budget : 10.000 configurations ● Model shifting : ○ From 6.2% to 6.7% error rate ● Scratch : ○ From 7.1% to 7.7% error rate ● Incremental Shifting : ○ From 6.1% to 6.7% 10 64
- 65. Results Budget : 1.000 configurations ● Model shifting : ○ From 6.7% to 10.6% error rate ● Scratch : ○ From 14.9% to 16.7% error rate ● Incremental Shifting : ○ From 6.7% to 13.3% Budget : 5.000 configurations ● Model shifting : ○ From 5.6% to 7.1% error rate ● Scratch : ○ From 8.2% to 9.2% error rate ● Incremental Shifting : ○ From 5.6% to 7.5% Budget : 10.000 configurations ● Model shifting : ○ From 5.2% to 6.1% error rate ● Scratch : ○ From 7.1% to 7.7% error rate ● Incremental Shifting : ○ From 5.2% to 6.5% Model 4.13 Budget : 85.000 configurations Model 4.13 Budget : 20.000 configurations Budget : 1.000 configurations ● Model shifting : ○ From 8.5% to 11.6% error rate ● Scratch : ○ From 14.9% to 16.7% error rate ● Incremental Shifting : ○ From 8.5% to 13.8% Budget : 5.000 configurations ● Model shifting : ○ From 6.7% to 7.9% error rate ● Scratch : ○ From 8.2% to 9.2% error rate ● Incremental Shifting : ○ From 6.7% to 7.9% Budget : 10.000 configurations ● Model shifting : ○ From 6.2% to 6.7% error rate ● Scratch : ○ From 7.1% to 7.7% error rate ● Incremental Shifting : ○ From 6.1% to 6.7% 10 65
- 66. Summary ● Model 4.13 is saved ○ Positively reuse old model on new version at lower cost ○ Better than learning from scratch for years ● Incremental Shifting ○ More sensible to previous models error ○ Better use of more transfer budget 11 66
- 67. Kpredict Python module for Python 3.8+ ( https://github.com/HugoJPMartin/kpredict ) Error : ≃ 6.3% 97% of the predictions are below 20% error 12 67
- 68. 68