Engineering Design Principles

Clearly defined problem
- Assess the efficacy of various denoising filters in preserving the accuracy of image classifier models under a noise-based attack.
Requirements
- Only algorithmic approach for defense
- Must be faster than auto-encoder
Constraints
- Computing power
- Memory usage
- Impossible to know who and how a model will be attacked
Engineering standards
- PEP 8
Cite applicable references
Considered alternatives a) Iterate on the design i) Advantages - Potentially more computationally efficient than an ML approach - Will likely use less memory than a model used to clean inputs - No training (very computationally intense) stage ii) Disadvantages - Potentially less effective than than an ML approach iii) Risks - Conventional algorithm may be more vulnerable to reverse engineering
Evaluation process
- Cross validation
- Effectiveness will be measured as the percent of correct classifications
- Testing clean vs. filtered training data
- Ablation variables:
  - Different models
  - Different datasets
  - Different filters
Deliverables and timeline