Orateur
Description
Dimensionality reduction techniques, such as PCA, t-SNE, and UMAP, are commonly used in various scientific fields to analyze high-dimensional data. These methods project the data onto a lower-dimensional embedding space, making it easier to visualize and analyze. In this conference presentation, I will discuss two different aspects of dimensionality reduction. First, I will present an application of non-linear projection methods in the field of neuroscience, specifically focusing on the spatial representation in the hippocampus of mice. I will highlight how these techniques can reveal the spatial coding modalities and mechanisms at play. Second, I will present analytic results for non-linear projection of paradigmatic data, which can serve as benchmarks for studying real-world data. These benchmark results can provide a reference point for evaluating the performance and effectiveness of different dimensionality reduction techniques in practical applications.
Affiliation de l'auteur principal | INMED, Aix-Marseille Université |
---|