This summer, I had the incredible opportunity to work on a project titled "Contextualizing the Accuracy-Fairness Trade-off in Algorithmic Prediction Outcomes" as part of the Lehigh STEM-SI program. I had the honor of presenting this work at the Creative Inquiry@Lehigh University Mountaintop Summer Expo last Friday.
Throughout this project, I delved into the intricate balance between accuracy and fairness in AI systems, exploring how different conditions like algorithmic confidence, label errors, and sample sizes impact this trade-off. The insights gained from this research are pivotal in understanding how to design more equitable AI systems. I enhanced my ability to manipulate and analyze large datasets, ensuring robustness and fairness, and gained proficiency in developing and adjusting algorithms to prioritize fairness without compromising accuracy. Additionally, I deepened my understanding of the ethical implications of AI and the importance of transparency and accountability in AI systems, and improved my ability to communicate complex ideas effectively, both in written and oral formats.
These skills are not just valuable academically but are crucial for my future career in data science and AI. I am excited to apply these insights to create more fair and effective AI solutions in real-world applications. A heartfelt thank you to my project mentor, Professor Kofi Arhin, for his guidance and support throughout this journey. His expertise and encouragement were instrumental in the success of this project.
#LehighUniversity#STEMSI#AI#DataScience#EthicalAI#ResearchExperience#FairnessInAI#CareerGrowth
Artificial intelligence is revolutionizing research. Although one school of thought still questions the ethical use of AI, especially in academic research. ☯️
There have been studies which highlight the use of AI tools that not only help in identifying articles for literature review, extracting relevant information and synthesizing findings but also in structuring manuscripts, providing clarity while adhering to academic standards. 🌟💻✅
This open house provides a fantastic opportunity to connect with faculty members and research enthusiasts to understand how collaboration with AI can help in saving time energy and efforts by focussing on analysis and innovation rather than getting bogged down by intricacies of language. 👩🔬👨🔬👨🏫👩🏫
Whether you're considering pursuing a PhD or simply interested in the intersection of AI and academic research, we welcome you to join us! 🙏
Do register through the link mentioned in the post below, looking forward to seeing you here at the lovely JGU campus in Sonipat. (The Institute of Eminence).
#jgu#ods#instituteofeminance#doctoralstudies#academicresearch#officeofdoctoralstudies
Curious about how to prepare students to thrive in the era of AI? The #IGEHub project led by Misha Chertkov at University of Arizona integrates data science and AI into its Applied Mathematics PhD program.
Collaborating with national and industrial labs, it equips a new cohort of researchers to tackle pressing global issues. The initiative includes modernizing the AM curriculum to meet national security needs and forming triads (student, university advisor, lab researcher) to offer PhD students broader career paths, especially in non-academic sectors. It aims to attract diverse talent and create a workforce ready for the AI-driven landscape, establishing a replicable model for other universities.
Learn more at the IGE Hub: https://bit.ly/3JytnFO
The IGE Hub is generously supported by the National Science Foundation (NSF).
I just wanted to take a moment to express my deep appreciation for our programme director and professor Esther Mondragón . I've had the privilege of being a student in her Cognitive Computational Model course this term, and it's been an enlightening experience.
I have completed several courses related to AI, but never learned AI from a psychological perspective. At first this module appeared to be uninteresting, as it had no coding and AI concepts I was accustomed to, but it turned out to be the most relevant and up-to-date content I was missing.
From intricacies of Turing Machines, Marr's levels of analysis, Pavlovian Conditioning, Bayesian Inference, Church-Turing thesis, to every other concept taught was meticulously prepared. Indeed, there is no way to grow other then familarising yourself from foundations.
"Foundational roots are growth's solitary origin"
Esther is excellent at her work and profound in her field. Significantly, when I googled to research more, the top results included research paper and sites published by her.😃
Irrespective of my coursework outcome 😄, I enjoyed this module and its content completely. I am glad I chose this course.
Thank you, Esther, for a memorable term. Your way of teaching made this toughest module of the term much more easier, interesting and entertaining. #appreciationpost#MScAI#cognitivecomputing
🎓 Excited to share that I’ve recently started my postgraduate degree at UCL, pursuing a master’s in AI for Sustainable Development! 🌍🤖
Coming from an Applied Maths background , I had the chance to explore the mathematics behind AI & ML during my undergraduate studies. That foundational knowledge sparked a curiosity in me, and I knew I wanted to dive deeper into these technologies at a postgraduate level.
I’m really excited about expanding my understanding of AI, especially at the intersection of sustainability and technology—two areas I’m deeply passionate about. I look forward to learning more and collaborating with others who share the vision of using AI for a better, more sustainable future!
#AI#SustainableDevelopment#UCL#PostgraduateLife#TechForGood#AppliedMaths#MachineLearning
Data Science Team Leadership | Psychometrics | Machine Learning | AI | Statistics | Time Series/Forecasting | Python | R | Shiny | SQL | Data Visualization. Clear stories from complex data.
Thankyou very much Cian Byrne for inviting me to talk to your students. It was a real pleasure. Following on from the talk, what I would really like to stress to those embarking upon a career in data science, Artificial Intelligence/Machine Learning or statistics is the following:
1) Social and legacy media are full of "AI takes" (my LinkedIn feed seems to be jammed with these lately). Take them with a (big) grain of salt, particularly those about Large Language Models and their capabilities.
2) Try to solve a real business problem with AI/ML; and communicate as best you can with the business and Subject Matter Experts. When David Stillwell and I built NESA's HSC Minimum Standards Computerised Adaptive Tests for Reading and Numeracy, we solved a long running business "sore", increased efficiency and built capability. In fact, you should make solving a business problem, or attempting to, your first order of business. Don't be shy, most SMEs really respond well to people who are both genuinely interested in what they do and are trying to help them.
3) You might have heard that 90 - 95% of the work in an AI/ML project concerns the sourcing, cleaning and preparation of data. This is actually true. Most organisations do not have their datasets ready to use for AI because they have been prepared to serve business needs and requirements. You are going to spend a lot of time getting your hands dirty with good old-fashioned Extraction, Transformation and Loading. Get a good feel for your data, you may start noticing small but important things.
4) The real world of employment will throw all kinds of problems your way to solve, so don't limit your skill set to Deep Learning or XGBoost because they work well on Kaggle. Learn the full suite of statistical tools from linear regression to support vector machines. Statistical inference and null hypothesis significance testing are important, as well as time series/forecasting and survival analysis. Oh yeah, developing some subject matter expertise in your industry of choice will not be a bad thing either.
5) Remember the basics and always evaluate your models on a test dataset or via cross-validation. Start with a high bias/low variance model first to establish a baseline. Overfitting your data in an attempt to get the highest F1 or accuracy score is bad. Don't be afraid to call out problems or limitations with the data, this mitigates risk.
6) Avoid the whole R versus Python (versus Rust vs whatever) thing and other pointless debates. Learn both and play each to their strengths. SQL is a must.
7) Carefully build a network of professional contacts over time. Your fellow data scientists will help you keep it real and help bring opportunities to your attention.
Happy for others to fill in what I have (most likely) missed.
University of Nebraska Board of Regents Approves Budget and Launches AI Degree
Exciting developments in the academic world as the University of Nebraska's Board of Regents has given the green light to the 2025-2027 budget request. This significant milestone is set to advance higher education funding as it moves to the Coordinating Commission for Postsecondary Education and, subsequently, the state for final approval.
A highlight of this budget is the introduction of a new degree program in artificial intelligence. This forward-thinking initiative aims to equip students with cutting-edge skills in AI, preparing them for the evolving job market and innovation landscape.
Stay tuned for more updates as the University of Nebraska continues to enhance its academic offerings and contribute to the field of artificial intelligence.
#HigherEducation#ArtificialIntelligence#UniversityOfNebraska#Innovation#AcademicExcellence#FutureSkills
What's up, Prof? I think most people have just a rough idea of what a professor 👨🏫 does all day. That's why I've decided to post about what my week looks like from time to time this semester. Here we go for week 41:
⚡ Flashback: We kicked off the winter semester last week, and the university is crowded again with students. My lectures start today, and I am very excited to meet the new group of students in the AI master program.
📅 Monday: Lectures in Advanced Computer Vision and Applied Deep Learning in the AI master, some meetings to discuss laboratory equipment.
📅 Tuesday: Lectures in ACV and ADL again, faculty meeting, meeting with my BA and MA students, and meeting of the doctoral program committee.
📅 Wednesday: Working on research proposals, preparing lectures for next week
📅 Thursday: Meetings with all my PhD students. CVIMS Lunch
📅 Friday: Welcome Day at THI Campus für Weiterbildung for my new Applied AI master students
💡 Highlight of the week: My winter semester modules start this week, after 3 months without lectures, I'm looking forward to finally standing in front of the students again 👨🏫
🍬 Bonus: You know what those What's up, Prof posts need? Memes, that's what 😎!
#whatsupprof#professor#artificialintelligence#machinelearning#lecture#researchTechnische Hochschule Ingolstadt
