0% found this document useful (0 votes)
10 views3 pages

ML Assignment 8

The document outlines a Machine Learning assignment focusing on Ethical AI, Causal Inference, and TinyML, with objectives including understanding cause-and-effect through Structural Causal Models, designing fair systems, deploying models on low-power devices, and securing models against adversarial attacks. It covers various techniques and case studies, such as fairness metrics, pruning methods, and federated learning with differential privacy. Submission requirements include a technical report, code repository, and a reflection on the 'Right to Explanation' in automated decision-making.

Uploaded by

xahid hasan
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
10 views3 pages

ML Assignment 8

The document outlines a Machine Learning assignment focusing on Ethical AI, Causal Inference, and TinyML, with objectives including understanding cause-and-effect through Structural Causal Models, designing fair systems, deploying models on low-power devices, and securing models against adversarial attacks. It covers various techniques and case studies, such as fairness metrics, pruning methods, and federated learning with differential privacy. Submission requirements include a technical report, code repository, and a reflection on the 'Right to Explanation' in automated decision-making.

Uploaded by

xahid hasan
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd

Machine Learning Assignment 8: Ethical AI, Causal Inference &

TinyML

1. Causal Inference and Directed Acyclic Graphs (DAGs)

Objective: Move beyond correlation to understand cause-and-effect.

 Structural Causal Models (SCM): Define the difference between


$P(Y | X)$ (observational) and $P(Y | do(X))$ (interventional).

 The Backdoor Criterion: Identify which variables must be


controlled for to block all spurious paths between a treatment
and an outcome.

 Implementation: Use the DoWhy library to estimate the causal


effect of a feature on a target variable in a dataset with known
confounders.

2. Algorithmic Fairness and Bias Mitigation

Objective: Design systems that do not perpetuate societal prejudice.

 Fairness Metrics: Define and calculate "Equalized Odds,"


"Demographic Parity," and "Counterfactual Fairness" for a loan
approval model.

 Mitigation Strategies: Implement a Pre-processing (Re-


weighing), In-processing (Adversarial Debiasing), and Post-
processing (Reject Option Classification) technique.

 Case Study: Analyze the "Gender Shades" study and explain


why intersectional analysis is critical for facial recognition
systems.

3. TinyML and On-Device Intelligence

Objective: Deploy ML models on ultra-low-power microcontrollers.

 Pruning and Sparsity: Explain how magnitude-based pruning


removes "unimportant" neurons to reduce model size without
significant accuracy loss.

 Knowledge Distillation: Train a "Teacher" model (Complex CNN)


and use it to guide the training of a "Student" model (TinyNet).

 Hardware Constraints: Profile a model’s RAM and Flash usage


for a target device like an Arduino Nano 33 BLE or ESP32.
4. Robustness against Adversarial Attacks

Objective: Secure models against malicious manipulation.

 Fast Gradient Sign Method (FGSM): Generate adversarial


perturbations to fool a pre-trained ImageNet model.

 Defensive Distillation: Explain how this technique can increase


the "smoothness" of the model's decision surface to resist
attacks.

 Certified Robustness: Discuss the concept of "Randomized


Smoothing" as a way to provide a mathematical guarantee on
model stability.

5. Federated Learning with Differential Privacy

Objective: Secure collaborative learning across edge devices.

 Secure Aggregation: Describe the protocol that allows a


central server to sum local model updates without ever seeing
individual device data.

 The Privacy Budget ($\epsilon$): Track the cumulative privacy


loss over multiple communication rounds of Federated
Averaging (FedAvg).

 Implementation: Use PySyft or Flower to simulate a federated


training environment across five virtual clients.

6. Symbolic AI and Neuro-Symbolic Integration

Objective: Combine deep learning with logic-based reasoning.

 Logic Tensor Networks (LTN): Explain how First-Order Logic


formulas can be grounded into neural network operations.

 Program Synthesis: Discuss how models like DeepCoder use ML


to predict which logical primitives are needed to solve a
programming task.

 Benefits: Why does adding a symbolic layer improve the


interpretability and "out-of-distribution" generalization of a
neural network?

7. Explainable AI for Computer Vision (XAI-CV)

Objective: Visualize the "Look" of a model's decision.


 Grad-CAM: Implement Gradient-weighted Class Activation
Mapping to produce heatmaps showing which pixels
contributed most to a specific classification.

 Influence Functions: Use influence functions to find the "Most


Influential" training images that led to a specific test-time
prediction.

 Human-Centered Evaluation: Conduct a small study to see if


these explanations actually help a non-expert trust the model's
decisions.

Submission Requirements

 Technical Report: 10 pages total. Must include DAG diagrams,


fairness metric tables, and Grad-CAM heatmaps.

 Code Repository: Python scripts using DoWhy, AIF360,


TensorFlow Lite, and Art-Toolbox.

 Reflection: A 1-page essay on the "Right to Explanation" in


automated decision-making as outlined in the GDPR.

You might also like