APPLIED AI #9 REPORT

The 9th Applied AI Study Group has come to an end - what a milestone it’s been, packed with exciting firsts! After a few years, the Applied AI Study group held on-site sessions on Saturdays, providing a hands-on learning environment for AI enthusiasts. During the week, we gathered in the evenings to discuss the solutions for the five homework notebooks.

Out of 38 applicants, we kicked off the program with 20 participants, and 15 dedicated members earned their place as successful graduates.

Under the leadership of our team member Onur Boyar, our guides and contributors crafted an exceptional learning experience. Huge thanks to Hasan Avcı, Ece Özen İldem, Karahan Şahin, Yağız İpekçi, Enes Sadi Uysal, Berkin Deniz Kahya, and Melih Darcan for the amazing curriculum and materials!

Take a look at the syllabus!

Plus, we organized the first AI Safety Talks event with AI Safety Türkiye and hosted Erdem Bıyık!

GUIDEs:

SYLLABUS OVERVIEW:

During the first two weeks, we explored how Machine Learning in Business works with the guidance of Hasan Avcı. The section began with an introduction to the role of a Machine Learning engineer, delving into responsibilities, skills, foundational tools and essential frameworks. The session progressed to converting the prototype into a functional ML product by constructing an API and integrating MLOps elements for scalability, reliability, and maintainability.

Complementing this theoretical knowledge, Ece Özen İldem led two lab sessions on ML in Business. This session reinforced the concepts learned through interactive exercises and real-world applications, providing deeper insights into effectively implementing machine learning solutions in business contexts.

During the sessions on Natural Language Processing, led by Karahan Şahin, we focused on both historical context and modern advances in NLP, from traditional text processing to the latest in transformer-based models. Participants studied core transformer concepts, including BERT and its derivatives (RoBERTa, ALBERT, DistilBERT), and learned to fine-tune models on benchmarks like GLUE and SuperGLUE. Generative models like GPT-2, GPT-3, and GPT-4 were introduced along with prompt engineering and in-context learning techniques.

In the third and fourth weeks, Yağız İpekçi led NLP Lab Sessions where participants got their hands dirty with practical activities and interactive exercises to deepen their understanding of the concepts.

During the week on Advanced Computer Vision, led by Enes Sadi Uysal with the assistance of Berkin Deniz Kahya and Melih Darcan, we delved into topics such as diffusion models, visual transformers, and multimodal approaches. We began with the basics of computer vision, then explored transformers on image data, diffusion models, and multimodal approaches such as CLIP. 

Berkin Deniz Kahya led a Computer Vision Lab Session, providing participants with practical, hands-on experience. This lab allowed attendees to apply the theoretical knowledge gained, enhancing their understanding through interactive exercises and real-world applications.

In the final week, Onur Boyar moderated an exciting series of talks by Gizem Tanrıver, Oğuz Kaplan and Elvan Karasu. Gizem Tanrıver shared her professional experience in the field of AI applications in healthcare, while Oğuz Kaplan covered machine learning in finance and risk assessment. Elvan Karasu presented her latest published article named “Temporal and Featurewise Attention for Alzheimer's Disease Conversion Prediction”. Onur Boyar shared his recent academic and industry studies on De Novo Design of Molecules and Crystals with Generative AI. After an open discussion, the course concluded with a graduation ceremony!

COMMUNITY FEEDBACK

 “The structure was well-organized, combining theory with practical assignments, live sessions, and labs. The mentors were highly knowledgeable and supportive throughout the journey. This program significantly strengthened my applied AI skills and prepared me not only with real world challenges but also enabled me to be more confident about my abilities in NLP and ML in not only industrial/professional but also in academic settings.“

“The instructors were amazing — they not only knew their stuff but also taught us the tools that are actually being used in the industry right now, which made everything feel so much more practical. On top of that, they were super approachable and made communication really easy. It felt like I was learning from people who genuinely cared about helping us succeed. Being a part of the inzva community was also a huge plus — I met so many motivated, like-minded people along the way.”

HOW DID OUR PARTICIPANTS DO?

Stay tuned for the next batch!

All participants have to abide by our CODE OF CONDUCT  and LETTER OF CONSENT. 

A  BEV Foundation project inzva is a non-profit hacker community organizing study and project groups as well as camps in the fields of AI and Algorithm; and gathering CS students, academics, and professionals in Turkey.