Artificial Intelligence for Goods: Cases and Tools
1.
2. About
1. CEO DevRain, devrain.com.
2. CTO ДонорUA, donor.ua.
3. PhD in Computer Science.
4. Microsoft Regional Director
5. Microsoft AI Most Valuable Professional
6. Open data, Smart City expert.
7. Ex-EGAP Challenge coordinator.
8. The Best Professional in Software Architecture
(Ukrainian IT Award).
3. Global issues
1. Employment and skills
2. Climate change
3. Food, Energy and Water
4. Health, Sleep, Nutrition
5. Media Biases and Access
6. Corruption
7. People with disabilities
8. Gender Equality
9. Ending Poverty
10. Ageing
11. AIDS
12. Human Rights
13. Saving animals
14. International Law and
Justice
15. Migration
16. Oceans and the Law of
the Sea
17. Peace and Security
18. Population
19. Refugees
20. Africa
21. Fakes
22. Privacy
23. Quality education
24. Democracy
25. Terrorism
26. Populism
Source: United Nations
4. How can AI and machine learning
be applied to solve some of
society's biggest challenges?
5. Process
1. Creating dataset (open data, manual, IoT)
2. Understanding data (feature engineering)
3. Model
1. Creation
2. Evaluation
3. Validation
4. Improvement
6. Types of problems
1. Clustering
1. Given news articles, cluster into different types of news.
2. Classification
1. Yes/No - binary classification
2. “Is this picture a cat or a dog or a tiger?” - multi-class
classification
3. Association analysis
1. If {A, B} then {C}
4. Regression
1. What is the price of house in a specific city?
7. CareerVillage.org
Nonprofit that crowdsources career advice for
underserved youth.
Your objective: develop a method to recommend
relevant questions to the professionals who are
most likely to answer them.
https://www.kaggle.com/c/data-science-for-good-careervillage
8. DonorUA: predicting if person
will donate a blood
Features:
1. Recency – months since last donation
2. Frequency – total number of donation
3. Monetary – total blood donated in c.c.
4. Time – months since first donation
Full article:
http://devrain.com/posts/solving-classification-
problems-with-azure-machine-learning-for-blood-
donations-prediction
10. DonorUA
1. Social media monitoring
2. Donor – recipient match
3. Predicting supply/demand
4. Donor’s portrait
http://blog.1991.center/donorua
Join our Telegram channel:
https://t.me/donorua
11. Language Understanding (LUIS.ai)
A machine learning-based service to build natural language into apps, bots, and IoT devices.
Quickly create enterprise-ready, custom models that continuously improve.
16. Fighting fire with machine learning:
using TensorFlow to predict wildfires
“We decided to develop a device that
could identify and predict areas in a
forest that are susceptible to wildfires,
providing an early warning to fire
departments.
Using TensorFlow, we can analyze images
of biomass and estimating their moisture
content and size to determine the
amount of dead fuel.”
https://www.blog.google/technology/ai/fighting-fire-machine-
learning-two-students-use-tensorflow-predict-wildfires/
17. AI helps farmers identify diseased plants
An example of a
diseased cassava leaf.
Cassava is a crop that
provides for over half
a billion people daily.
https://www.blog.google/technology/a
i/ai-takes-root-helping-farmers-
identity-diseased-plants/
18. Вирубка лісів
Згідно зі статистикою організації Глобального
моніторингу лісів, за останні 15 років Україна
втратила майже 500 тис. га лісу. З 2015 року,
коли у країні був введений мораторій на
експорт лісу-сировини, як не парадоксально
звучить, показники вирубки лісів масово
збільшилися. За оцінками деяки фахівців
протягом 2016 року в Україні було вирубано
16,4 млн кубометрів деревини, з них 8-9 млн
кубометрів легально.
http://texty.org.ua/d/deforestation-longread/
http://texty.org.ua/pg/blog/nartext/read/76201/J
ak_big_data_i_drony_mozhut_vratuvaty
19. Seeing AI
A free app that narrates
the world around you.
Designed for the low
vision community, this
research project
harnesses the power of
AI to describe people,
text and objects.
https://www.microsoft.com/en-
us/seeing-ai
20. Wild Me
Wild Me used computer vision and deep
learning algorithms to create a platform called
Wildbook, which scans millions of
crowdsourced wildlife images at scale.
Wildbook can identify the species as well as
the individual animal, and the public can follow
the movements of their favorite animals. The
aggregated data is used by scientists to help
inform conservation decisions. Microsoft is
supporting their efforts by hosting Wildbook
on Azure and making Wild Me’s open source
algorithms available as APIs.
https://www.microsoft.com/en-us/ai/ai-for-earth-
projects?activetab=pivot1:primaryr3
34. GPT-2
GPT-2 generates synthetic
text samples in response to
the model being primed with
an arbitrary input. The model
is chameleon-like — it adapts
to the style and content of the
conditioning text. This allows
the user to generate realistic
and coherent continuations
about a topic of their
choosing.
https://blog.openai.com/better-language-
models/#sample8
36. Programs and Organizations
• AI for Sustainable Global Development
https://ai4good.org/
• Google AI for Social Good
https://ai.google/social-good
• AI for good with Microsoft Artificial Intelligence
https://www.microsoft.com/en-us/ai/ai-for-good