SMLP tutorial

[mdmitry1]

This tutorial has SMLP 3 examples of various complexities:

Example	Variables	Objectives	Constraints	Difficulty
Eggholder	2	1	None	High (multi-modal)
DORA	2	1	1 (circular)	Low (convex)
BNH	2	2	2 (non-linear)	Medium

Each example contains:

1. Reference
2. Push button script
3. Vizualization

[mdmitry1]

Initial revision - please, review

[mdmitry1]

Added explicit mention that SMLP supports black-box functions

[mdmitry1]

1. Added black-box optimization applications reference to README.md
2. Added reference to black box function definition used in the README.md file
3. Added SMLP flow stages description
4. Multiple style improvements

[zurabksmlp]

The comments below are on README.md

1. This repository contains  This tutorial contains
2. Symbolic Machine Learning and Prediction  Symbolic Machine Learning Prover
3. constraints defining the set Ω is unknown -> what is set Ω here? Definition is missing?
4. Structure of the objective function of is unknown  the second “of” is redundant? Or it must be “f”?
5. 5. Regarding this paragraph:
      SMLP optimization flow is comprised of two stages:

• Model build: input data is converted into the one of 3 types of models:
  1. Polynomial model
  2. Decision tree
  3. Neural network model
• Optimization: model and constraints are used to find objective function(s) minimum considering input constraints

a. Model build: input data is converted into the one of 3 types of models:  There are more types of tree models supported in SMLP, not just Decision trees.
b. Besides Optimization, there are many more other modes as well like synthesis, verification, certification, querying, subgroup discovery, DOE, …. I thin you can say that this tutorial focuses on optimization (and will be extended in future to cover some other modes as well)
6. Maybe add SMLP results for the three examples? They can be found in thre results files but for the reader it will be more convenient to include in the text (so that one can compare with expected results).
7. Maybe add reference to SMLP tool paper: https://link.springer.com/chapter/10.1007/978-3-031-65627-9_11
Brauße, F., Khasidashvili, Z., Korovin, K. (2024). SMLP: Symbolic Machine Learning Prover. In: Gurfinkel, A., Ganesh, V. (eds) Computer Aided Verification. CAV 2024. Lecture Notes in Computer Science, vol 14681. Springer, Cham. https://doi.org/10.1007/978-3-031-65627-9_11
There is also a free ArXiv report version https://arxiv.org/abs/2402.01415

[konstantin-korovin]

Good suggestions Zurab!

…

On Tue, 10 Feb 2026 at 11:36, zurabksmlp ***@***.***> wrote: ***@***.**** commented on this pull request. The comments below are on README.md 1. This repository contains  This tutorial contains 2. Symbolic Machine Learning and Prediction  Symbolic Machine Learning Prover 3. constraints defining the set Ω is unknown -> what is set Ω here? Definition is missing? 4. Structure of the objective function of is unknown  the second “of” is redundant? Or it must be “f”? 5. 5. Regarding this paragraph: SMLP optimization flow is comprised of two stages: - Model build: input data is converted into the one of 3 types of models: 1. Polynomial model 2. Decision tree 3. Neural network model - Optimization: model and constraints are used to find objective function(s) minimum considering input constraints a. Model build: input data is converted into the one of 3 types of models:  There are more types of tree models supported in SMLP, not just Decision trees. b. Besides Optimization, there are many more other modes as well like synthesis, verification, certification, querying, subgroup discovery, DOE, …. I thin you can say that this tutorial focuses on optimization (and will be extended in future to cover some other modes as well) 6. Maybe add SMLP results for the three examples? They can be found in thre results files but for the reader it will be more convenient to include in the text (so that one can compare with expected results). 7. Maybe add reference to SMLP tool paper: https://link.springer.com/chapter/10.1007/978-3-031-65627-9_11 Brauße, F., Khasidashvili, Z., Korovin, K. (2024). SMLP: Symbolic Machine Learning Prover. In: Gurfinkel, A., Ganesh, V. (eds) Computer Aided Verification. CAV 2024. Lecture Notes in Computer Science, vol 14681. Springer, Cham. https://doi.org/10.1007/978-3-031-65627-9_11 There is also a free ArXiv report version https://arxiv.org/abs/2402.01415 — Reply to this email directly, view it on GitHub <#49 (review)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AIVYHTKLG3PRPLVXUJ3BMYL4LG7EJAVCNFSM6AAAAACUPTGQCSVHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHMZTONZYGQZDAMBQGI> . You are receiving this because your review was requested.Message ID: ***@***.***>

[mdmitry1]

Updated README.md after pull request review

[mdmitry1]

Merged with main branch and successfully reproduced results for all three tests in tutorial

[mdmitry1]

Added link to SMLP User Manual

[mdmitry1]

Added Intel Signal Integrity test

[mdmitry1]

Successfully reproduced results for Intel signal integrity test in Docker container

[mdmitry1]

Added instructions for running Intel test in Docker environment

[mdmitry1]

Improved benchmark by replacing Docker Desktop by Docker CLI

[mdmitry1]

Added plotting