Kristina P. Sinaga
"A mind wired for patterns. A soul allergic to nonsense."
About Me
As an independent researcher, I specialize in applied mathematics, machine learning, and federated learning, driving advancements at the intersection of academia and industry. With a robust background as a postdoctoral researcher at ISTI-CNR, Italy and Department of Applied Mathematics, CYCU, Taiwan. I have a proven track record of developing innovative solutions that bridge theoretical research and practical applications, delivering impactful results in both domains.
Research Interests
My research focuses on developing innovative algorithms for complex systems, with a particular emphasis on multi-view clustering, heat-kernel methods, and fuzzy clustering approaches in both centralized and decentralized environments. These methods are designed to deliver robust performance in unpredictable, real-world settings characterized by heterogeneous data sources. For example, in applications like autonomous systems or smart healthcare, where data varies widely in format and quality, my algorithms ensure reliable outcomes despite inherent uncertainties.
Specializations
I focus on unpredictable, real-world settings where systems are truly tested. My approach blends theoretical rigor with hands-on experimentation to build scalable, intelligent solutions that hold up beyond the lab. My work with heat-kernel coefficients has led to novel clustering algorithms that outperform traditional methods on heterogeneous data.
Experience & Projects
Algorithm Design
The core of my work lies in creating scalable solutions that bridge theoretical rigor with practical utility. Over the past eight years, I have specialized in unsupervised learning, developing clustering algorithms that outperform traditional methods when handling diverse, multi-source data. This expertise provides a strong foundation for my current exploration into hybrid models that integrate unsupervised, supervised, and deep learning techniques. This transition represents a paradigm shift, leveraging the strengths of deep learning to enhance efficiency and accuracy in complex data environments, such as real-time analytics in IoT systems or personalized recommendation engines.
Robustness Research
A key aspect of this research involves adapting to modern computational tools. Transitioning from MATLAB to Python has been a significant evolution in my workflow, enabling greater flexibility and access to advanced libraries like TensorFlow and PyTorch. This shift is straightforward yet impactful: it begins with formulating core functions and progresses to leveraging Python’s extensive ecosystem of libraries, streamlining algorithm development and deployment. For instance, Python’s versatility supports rapid prototyping for industry applications while maintaining the precision required for academic research. This research has broad implications. For academia, it advances the theoretical understanding of clustering and learning paradigms. For industry, it offers practical, scalable solutions for data-driven decision-making in dynamic environments. For the general public, it paves the way for smarter, more reliable technologies that adapt to real-world complexities, from improved medical diagnostics to enhanced user experiences in digital platforms.
NeuralGlow.ai
Founded a satirical AI lab blending humor, critique, and product design to challenge hype culture and reimagine intelligent systems.
Recent Publications
Browse my complete publication list with full citations and impact metrics
Opportunities
Open to New Roles
I am currently on the global job market for research-driven or applied roles—starting October 2025—with teams that value clarity, autonomy, and creative rigor.
If your work is serious about outcomes (and allergic to nonsense too), let's talk.
Get in TouchContact
Feel free to reach out if you’re interested in collaboration, research opportunities, or just want to discuss interesting problems in AI and mathematics.
- Email: ndkristinaps25@gmail.com
- Previous Affiliation: ISTI-CNR, Italy
© 2025 Kristina P. Sinaga | Last updated: September 2025