Applications of Machine Learning

Pre-approved Topic List

1. Applications of deep learning and reinforcement learning to train an autonomous agent to solve video game dynamics.
2. Implementations of unsupervised learning to identify structure in social media data, for instance the structure of a group of followers of a popular Twitter account.
3. Sentiment analysis techniques that can be used to understand public sentiment towards a policy, individual, or company using open social media platforms.
4. Modeling or predicting transportation patterns such as flight delays and traffic flows.
5. Modeling economic trends and fluctuations in real estate or stock markets.
6. Developing explanative or predictive models for the scores and performance of sports teams.
7. Using large data sets from public polling, censuses, and election results to understand the political geography of a country and assess the fairness of electoral maps
8. Using open source data to gain novel insights into the creation of “smart cities” or to improve the quality of life for urban residents (for example, by better understanding gentrification, the effects of development projects, the impact of housing initiatives, or the effects of particular transportation infrastructure approaches)
9. Understanding and modeling the ways in which algorithms can generate bias or ways to improve “data fairness”, with applications in the criminal justice system, corporate hiring practices, and the health insurance industry
10. Applications of machine learning or automation in chemical synthesis
11. Applications of machine learning predicting chemical reactions
12. Designing and developing materials for CO2 capture and storage using Machine Learning
13. Predicting climate change impacts on crop yields and detecting climate change-induced drought.
14. Modeling or predicting the effects of climate change, such as extreme precipitation, wildfire risk, and the degradation of forests.
15. Data scaping techniques that can be used to create novel data sets on important socio-economic phenomena
16. Understanding the fundamentals of natural language processing as it applies to machine translation
17. Creating recommendation algorithms for personal entertainment such as movie, TV, or book suggestions
18. Machine Learning driven development of better materials (batteries, solar panels, etc.)
19. Overview of supervised and unsupervised machine learning algorithms with use cases in business (customer analysis, churn rate, etc.)
20. How do people react to recommender systems (e.g., Alexa, ChatGPT, Siri) making mistakes? How do mistakes — and the type of mistake — affect a user’s trust and perception of the system? How does a user’s trust evolve over time, and how rapidly do users learn the limitations of recommender systems? These systems are rarely 100% accurate, so when should they say “I don’t know”?
21. How can a robot or “AI” system be a teammate instead of a tool? What makes an effective teammate? How can autonomous agents perceive the intentions of their human teammates and support their partners’ goals?
22. How smart is “too smart”? How can robots and recommender systems apply what they think the user knows, to conduct high-order reasoning? Do people like systems capable of complex reasoning?
23. ML in Astronomy: Simulation of CMB (relic Big Bang radiation) to understand early conditions of the universe
24. ML in Astronomy: Using data from SDSS (an instrument) to analyze galaxy spectra and galactic formation
25. ML in Astronomy: Using Kepler data to understand exoplanet light curve behavior

Topics in Image Recognition

1. Developing convolutional neural networks to learn to scan images, with applications in image recognition
2. Detecting image tampering, with applications in combatting fake news
3. Identifying objects and/or places in images
4. Creating stylized art/visual style transfer for other applications
5. Lensless imaging and Computational photography
6. Identifying words or bird calls in audio recordings
7. Low-power or low-light machine vision
8. Neuromorphic or “biologically plausible” machine learning
9. Data clustering
10. Multi-dimensional scaling and visualization

Topics in Medical Applications of Machine Learning

1. Utilizing predictive machine learning models to learn more about cardiovascular diseases such as stroke and heart disease
2. Utilizing predictive machine learning models to learn more about cancer prognosis and diagnosis
3. Applications of machine learning in modeling the spread of infectious diseases
4. Medical applications of machine learning training a ‘convolutional neural network’ to (for example, to predict skin lesions which are either benign or indicative of skin cancer)
5. Applications of Natural Language Processing in the health sector
6. A review on computer-aided drug design and discovery
7. Data and machine learning driven drug discovery with a case study on cancer, covid, or other diseases.