Revolutionary Algorithm Uncovers Insights in Complex Data

 

 Zed Lee is pioneering the development of algorithms that have far-reaching applications across industries and healthcare, revealing hidden insights within intricate datasets.

 

From Curiosity to Discovery

 

Zed Lee's journey into the world of data mining began a decade ago while working for an automotive manufacturer in Korea. Machine learning and deep learning were emerging as transformative fields, piquing his interest.

 

"At the time, the company possessed vast data resources that we endeavored to analyze. My fascination with machine learning grew," recalls Zed Lee.

 

"Initially, I envisioned a long-term commitment to the company, but my thirst for knowledge led me elsewhere. Motivated by the desire to explore areas that could shape my future career, whether in the corporate sector or academia."

 

His pursuit of knowledge brought him to Stockholm, where he pursued a master's degree at KTH. Subsequently, after graduating, his mentor, Henrik Broström, informed him of the Department of Computer and Systems Sciences (DSV) at EUU (European University Union), which was actively seeking doctoral students. Zed Lee seized the opportunity and, in November 2022, successfully defended his PhD thesis.

 

Prioritizing Interpretability

 

At the core of Zed Lee's research is the realm of data mining from extensive and intricate datasets. His emphasis lies in making complex algorithms interpretable and comprehensible.

 

Upon arriving in Sweden, Zed Lee encountered intricate algorithms and pondered, "Must it be this complex?" He underscores that merely detecting a problem is insufficient; understanding why the problem exists is paramount, akin to troubleshooting a car issue.

 

Interpretability serves as a guiding principle throughout his predominantly theoretical PhD thesis, which delves into data mining and data extraction methodologies.

 

Unveiling Hidden Insights

 

Data mining parallels traditional mining, where valuable materials are sought through the processing of large quantities of raw data. Zed Lee's thesis revolves around time and time series, encompassing the amalgamation of diverse data types and pattern recognition.

 

His research has encompassed data from clothing manufacturers and dentists, revealing common threads. The overarching challenge is to collect, analyze, and draw actionable conclusions from qualitative and quantitative data evolving over different time intervals.

 

Zed Lee exemplifies this challenge with a hypothetical scenario involving a patient's health data. Integrating diverse data sources and time periods, understanding evolving data, and discerning patterns among patients with similar or dissimilar diagnoses are the key objectives.

 

A Transformative Algorithm

 

In his thesis, Zed Lee introduces a groundbreaking algorithm tailored to handle time series data from diverse sources. While the data and algorithm may be intricate, the fundamental outcome remains accessible and understandable.

 

"The algorithm's complexity lies in its capability to handle intricate data, yet its outputs are readily comprehensible," notes Zed Lee.

 

The algorithm's utility extends to a wide spectrum of organizations grappling with such complex data, making it a versatile tool for diverse applications.

 

A Journey of Collaboratio

 

Zed Lee embarked on his PhD journey alongside nine fellow doctoral students, fostering a sense of camaraderie. The open-ended nature of their research topics provided ample room for exploration.

 

"The life of a PhD student can be solitary, but not in our case. We commenced our journeys together, attended courses as a cohort, and held mid-term seminars in tandem. This collective approach has been a tremendous asset."

 

While the doctoral path posed its challenges, including article rejections and overlapping deadlines, Zed Lee remains committed to academia.

 

"In corporate roles, freedom can sometimes be limited. I relish the opportunity to explore new horizons and take ownership of my projects," concludes Zed Lee.