Project brief
From scenario and baseline to reproducible evidence.
Use machine learning and experimental data to find optical-module settings that satisfy bit-error, power, temperature, and stability constraints.
The handbook starts with a runnable minimum, then adds real constraints. The final presentation should explain the method, results, failures, limitations, and next steps.
Evaluation focus
Use metrics to make trade-offs visible.
Project handbook
Project brief
Build a data-driven tuning system. Given module state and candidate parameters, predict outcomes and recommend the next experiment. Begin with a simulator or instructor-provided historical measurements.
Minimum version
Define a parameter table for bias, modulation, equalization, and temperature, with at least three measured outputs such as BER or Q-factor, power, throughput, or eye quality.
- A grid-search, random-search, or rule-based baseline.
- A surrogate model such as a random forest, Gaussian process, or small neural network.
- A single- or multi-objective parameter recommender.
- An explicit check that rejects unsafe or out-of-range suggestions.
Evaluation and boundaries
Use offline data or a constrained simulator. Students must not connect a model directly to a real laser, power supply, or communications link for automatic tuning. Report device, temperature, and constraint assumptions.
Resource note: External code, models, datasets, and platforms remain subject to their own licenses, terms of service, and applicable law.