#  Fall 2022 Lectures 

 



 ##  

  expand\_more  

 
  

 

### AI4SI book is available at [ai4sibook.org](https://ai4sibook.org)

SortDateTopicClusterReadings8/31

Introduction, class exercise: defining AI4SI in relation to other terms

**(**[**slides**](https://www.dropbox.com/s/mqt2hzqyz0oxq8n/Lecture%201.pdf?dl=0)**)**



 

- [AI4SI Part I: Introduction](https://ai4sibook.org/wp-content/uploads/2022/08/Introduction_to_AI_for_Social_Impact.pdf)
- [AI for Social Impact: Learning and Planning in the Data-to-Deployment Pipeline](https://teamcore.seas.harvard.edu/files/teamcore/files/ai_magazine_article.pdf)



9/7

Public Health: 1. Maternal and Child Health and 2. HIV Prevention

**(**[**slides**](https://www.dropbox.com/s/8w94fpjzvxmgywf/Lecture%202.pptx?dl=0)**)**

**(**[**Lecture video**](https://www.youtube.com/watch?v=0sAb_Xw9mS4) **of KDD 2022 keynote covering similar material)**



Case studies

Case studies readings:

1. [**Field Study in Deploying Restless Multi-Armed Bandits: Assisting Non-Profits in Improving Maternal and Child Health**](https://arxiv.org/abs/2109.08075)
2. [**AI-Augmented Interventions for HIV Prevention in Youth Experiencing Homelessness**](https://arxiv.org/abs/2009.09559)

**Additional background readings (optional):**

- [Restless Multi-Armed bandits (RMAB)](https://teamcore.seas.harvard.edu/files/teamcore/files/2016_15_teamcore_aamas2016_eve_yundi.pdf). Although presented on a different domain, this method is used in the maternal and child health case study (reading 1)
- [Social Networks](https://teamcore.seas.harvard.edu/files/teamcore/files/2016_10_teamcore_amulya_aamas16.pdf). Recommended to read first 4-5 sections to get further background for reading 2. Do not need to read the rest.



9/12

**Case studies presentations:** 1. Public Health and 2. Agriculture and Food Security

Case studies

1\. <a class="hwp-link">Deploying COVID-19 Case Forecasting Models in the Developing World</a>

2\. [Reinforcement Learning for Public Health: Targeted COVID-19 screening](https://ai4sibook.org/targeted-covid-19-screening)

3\. [Kudu: An electronic agricultural marketplace in Uganda](https://ai4sibook.org/kudu)

4\. [AI for Food Rescue](https://ai4sibook.org/food-rescue)



9/14

**Case studies presentations:** 1. Environment and Conservation and 2. Crisis Management and Disaster Response

Case studies

1\. [Poaching Risk Prediction and Multimodel Patrol Planning for Anti-Poaching](https://ai4sibook.org/anti-poaching)

2\. [The MegaDetector: Large-scale deployment of computer vision for conservation and biodiversity monitoring](https://ai4sibook.org/megadetector-conservation-biodiversity-monitoring)

3\. [The Google Flood Forecasting Initiative: Flood forecasting, crisis management &amp; disaster Response](https://ai4sibook.org/google-flood-forecasting-initiative)



9/19

Part 1: Participatory Approach Towards AI4SG. Invited lecturer: *Lily Xu, PhD Student, Harvard University*

Part 2: Case studies presentations



Case studies

- Envisioning Communities: A Participatory Approach Towards AI for Social Good \[[Harvard access](https://hollis.harvard.edu/permalink/f/1mdq5o5/TN_cdi_arxiv_primary_2105_01774)\] \[[arXiv](/cs288/Envisioning%20Communities:%20A%20Participatory%20Approach%20Towards%20AI%20for%20Social%20Good)\]

- [ADVISER: AI-Driven Vaccination Intervention Optimiser for Increasing Vaccine Uptake in Nigeria](https://www.ijcai.org/proceedings/2022/0712.pdf)
- [Using Machine Learning to Help Vulnerable Tenants in New York City](http://www.rayidghani.com/wp-content/uploads/2020/09/3314344.3332484.pdf)
- [ReforesTree: A Dataset for Estimating Tropical Forest Carbon Stock with Deep Learning and Aerial Imagery](https://ojs.aaai.org/index.php/AAAI/article/view/21471)
- [A GNN-RNN Approach for Harnessing Geospatial and Temporal Information: Application to Crop Yield Prediction](https://arxiv.org/abs/2111.08900) (9/21)



9/21

*Prof. Andrew Davies, Assistant Professor of Organismic and Evolutionary Biology, Harvard EOB*

**(**[**slides**](https://docs.google.com/presentation/d/1OJGjjay_gSLerNwb760izH-vQaTMmaHK/edit?usp=sharing&ouid=101984839737275208164&rtpof=true&sd=true)**)**



Project ideas

 

9/26

Quantifying bias in human and machine decisions. *Prof. Sharad Goel, Harvard Kennedy School*

 **(**[**slides**](https://docs.google.com/presentation/d/1t86dg2p0jIly5n8hVs-Es2CJA20jBlC6XTrWCWjkEPo/edit?usp=sharing)**)**



Project ideas

- [A large-scale analysis of racial disparities in police stops across the United States](https://5harad.com/papers/100M-stops.pdf)
- [Algorithmic Decision Making and the Cost of Fairness](https://5harad.com/papers/fairness.pdf)



9/28

*Prof. Chris Golden, Assistant Professor of Planetary Health and Nutrition at the Harvard TH Chan School of Public Health*

([slides](https://docs.google.com/presentation/d/1rNomdDewRiLmu6h95Lbr0QNqgbjWYDoV/edit?usp=sharing&ouid=110460147826990836818&rtpof=true&sd=true))



Project ideas

 

10/3

Embedded EthiCS by Michael Pope

**(**[**slides**](/file_url/133)**)**



Project ideas

- [AI Fairness Isn’t Just an Ethical Issue](https://hbr.org/2020/10/ai-fairness-isnt-just-an-ethical-issue)
- [Can you make AI fairer than a judge?](https://www.technologyreview.com/2019/10/17/75285/ai-fairer-than-judge-criminal-risk-assessment-algorithm/)



10/5

In-class OHs for class project

Project ideas

 

10/10

*No class – Indigenous Peoples' Day*

 

 

10/12

Public Safety: Stackelberg Security Games and their Applications

[**(slides**](/file_url/132)**)**

**(**[**Lecture video IJCAI 2018 keynote**](https://www.youtube.com/watch?v=O2su1u2AXG0) **covering similar material)**



Case studies

- [Stackelberg Security Games: Looking Beyond a Decade of Success](https://teamcore.seas.harvard.edu/files/teamcore/files/2018_11_teamcore_camera-ready.pdf)
- Assessing the Benefits and Costs of Homeland Security Research: A Risk-Informed Methodology with Applications for the U.S. Coast Guard ([Available on Canvas](https://canvas.harvard.edu/courses/106927/files/folder/Oct%2012%20readings?preview=15942862))
- Retrospective Benefit–Cost Analysis of  
    Security-Enhancing and Cost-Saving  
    Technologies ([Available on Canvas](https://canvas.harvard.edu/courses/106927/files/folder/Oct%2012%20readings?preview=15942878))



10/17

Self-selected paper presentations

Student presentations

- [Placement Optimization in Refugee Resettlement](https://pubsonline.informs.org/doi/epdf/10.1287/opre.2020.2093)
- [Efficient poverty mapping from high resolution remote sensing images](https://ojs.aaai.org/index.php/AAAI/article/view/16072)
- [Towards Facilitating Empathic Conversations in Online Mental Health Support: A Reinforcement Learning Approach](https://arxiv.org/pdf/2101.07714.pdf)
- [Mitigating Political Bias in Language Models Through Reinforced Calibration](https://arxiv.org/abs/2104.14795)
- [Learning Augmented Methods for Matching: Improving Invasive Species Management and Urban Mobility](https://ojs.aaai.org/index.php/AAAI/article/view/17727)



10/19

Self-selected paper presentations

Student presentations

- [Prediction of Landfall Intensity, Location, and Time of a Tropical Cyclone](https://arxiv.org/pdf/2103.16180.pdf)
- [FairFoody](https://ojs.aaai.org/index.php/AAAI/article/view/21447)
- [Why Is My Classifier Discriminatory?](https://proceedings.neurips.cc/paper/2018/file/1f1baa5b8edac74eb4eaa329f14a0361-Paper.pdf)
- [BeFair: Addressing Fairness in the Banking Sector](https://arxiv.org/pdf/2102.02137.pdf)
- [Bandit Data-driven Optimization for Crowdsourcing Food Rescue Platforms](https://www.aaai.org/AAAI22Papers/AISI-1095.ShiZ.pdf)



10/24

**Project progress presentations**

Student presentations

 

10/26

World Wildlife Fund

Case studies

 

10/31

Harvard Business School case: PAWS

Case studies

<https://teamcore.seas.harvard.edu/ai-conservation>

11/2



Implementation Science. *Shoba Ramanadhan, Assistant Professor of Social and Behavioral Sciences at the Harvard TH Chan School of Public Health*

**(**[**slides**](https://drive.google.com/file/d/1ADbn8_o1LGVgG1FsgYRkMrIQHJhADkfo/view?usp=share_link)**)**



Methods lectures

- [EPIS Framework website](https://episframework.com/)
- [Implementation science: What is it and why should I care?](https://www.sciencedirect.com/science/article/pii/S016517811930602X)



11/7

Lessons from ICT and International Development

Methods lectures

[AI4SI Part 3: Lessons from ICT and International Development](https://ai4sibook.org/wp-content/uploads/2022/08/lessons_from_ICTD.pdf)

11/9

Part 1: Causal Inference intro *Mateo Dulce Rubio, PhD candidate, Carnegie Mellon University*  
Part 2: RCT *Prof. Edward Kennedy, Associate Professor Department of Statistics &amp; Data Science, Carnegie Mellon University*

Methods lectures

[AI4SI Part 3: Randomized Experiments](https://ai4sibook.org/wp-content/uploads/2022/08/randomized-experiments.pdf)

11/14

Experimental Design. *Benton Taylor*  
*Assistant Professor of Organismic and Evolutionary Biology, Harvard University*

(slides in Canvas Files)



Methods lectures

[AI4SI Part 3: Applying Classic Concepts of Experimental Design in the Age of AI](https://ai4sibook.org/wp-content/uploads/2022/08/Experimental-Design.pdf)

11/16

HCI and AI4SG. *Herman Saksono, Assistant Professor, Northeastern University* 

Methods lectures

[AI4SI Part 3: User-centered Design in AI for Social Impact](https://ai4sibook.org/wp-content/uploads/2022/08/user-centered-design.pdf)

11/21

Project office hours

 

 

11/23

*No class – day before Thanksgiving*

 

 

11/28 

Final project presentations

 

 

11/30

Final project presentations