(99% of this post was generated)
In this episode of the DSB Podcast, I had the pleasure of sitting down with Ondra Draganov, a postdoc at INRIA in France, who has been a friend and a colleague for over 15 years. Ondra brings a wealth of knowledge in mathematics, having completed his bachelor’s and master’s in general mathematics and mathematical structures at Matfyz in Prague and later pursuing a PhD at IST Austria.
From Academia to Industry
Ondra shares his academic journey, detailing his recent transition to INRIA, a French institute focused on mathematics and computer science. Before this, he was in Vienna, working on his PhD at IST Austria, where he explored topological data analysis (TDA). TDA lie at the intersection of mathematics and computer science. Despite the theoretical nature of his work, Ondra finds motivation in the potential applications of his research, such as in biomedicine for cancer analysis.
Understanding Topological Data Analysis
Topological Data Analysis (TDA) may sound complex, but Ondra breaks it down into understandable terms. TDA is not a cure-all for data analysis but can be a powerful tool in the right context. It studies the ‘shape’ of data by analyzing the structure and connectivity of data points.
Ondra describes TDA as a method of examining the shape of data, which can be particularly useful when dealing with high-dimensional data where visual inspection isn’t feasible. For example, in time series data, TDA can identify periodic patterns that might signify meaningful events or states.
Practical Applications and Challenges
One of the intriguing discussions we had was about the practical applications of TDA. Ondra explains how TDA can be used in various domains, such as mobility data during COVID-19 restrictions or even in industrial settings for optimizing production processes. Although TDA provides features for further analysis, the real challenge lies in interpreting these features meaningfully.
Ondra also shared a fascinating application of TDA in linguistics, where it was used to compare embeddings of different languages, revealing insights into their structural differences.
Conclusion
As we navigate the complexities of data analysis, tools like TDA offer a unique perspective, allowing us to uncover patterns and insights that might otherwise remain hidden. Ondra’s insights into the practical applications and challenges of TDA provide a valuable guide for both newcomers and experienced practitioners looking to expand their analytical toolkit.
Links & Resources
Be First to Comment