Hi there! This is indeed a very interesting problem and similar to my work on second-hand automobile worthiness and classification algorithm using decision trees, t-SNE for dimensionality reduction, and the mixture of Gaussians. Here, I would like to mention that I am a researcher (Ph.D. candidate in mechanical engineering from METU, Turkey) myself and I have a strong aptitude towards data science and machine learning. As far as I understand (from the presentation), you require proposing a new method to write a paper through ensemble of the methods you mentioned. In this regard, I distinguish myself from the people who follow basic tutorials in sklearn or pytorch or tensorflow type of hack-the-ai tools. I provide the following skill-set, instead:
1- Strong academic writing skills
2- Advanced level of linear algebra, multivariable calculus, tensor mechanics, robotics
3- I can read and criticize research papers
4- MATLAB, Mathematica, C/C++, Python (and don't want to mention the ones I know but not use regularly)
5- Familiarity with fast linear algebra algorithms (e.g., BLAS), toolchain (e.g., Boost, Matlab prototyping), parallelization techniques (Algorithm conversion through splitting matrices, handling sparsity, etc)
NOTE: Ethical concerns (I won't accept jobs entailing plagiarism, college homework, university term projects, writing academic papers for someone else, so on, and so forth), instead I can provide collaboration and implementation.