#A brief description of my thesis that I’m writing
My thesis is about designing a Generative Adversarial Network architecture for compositional data.
Compositional Data
Compositional data is noteworthy because of this kind of data occurs in different contexts. Mathematically, compositional data is represented by points on a simplex , which means that given a point $$P_s(x_0,x_1,…,x_n)$$ then $$\sum_{k=0}^{k=n} x_k = 1$$
Some examples of compositional data are:
- Molar concentrations in chemistry.
- Rock composition in geology.
- Soil composition.
- DNA sequencing data
- Microbiome data. In particular, this is the data I’m working on, besides some toy models that come from probabilities distributions.