I'm really excited to share with you something I've been working on for a past few weeks. The paper is about mathematical model underlying PCA algorithm and it's done in online LaTeX editor, ShareLaTeX.
In Serbia, in order to graduate from high school (in USA that would be preparatory school I guess), among some other exams, you're obliged to write a final paper (graduation work) about a topic that's related to one of the subjects you studied at school. Although it's usually just a formality and it's generally not taken seriously, I, who had been exploring and studying ML and Deep Learning online (mostly at Coursera, but on Brilliant as well) during the last year, decided to put an effort to collect all the proofs and theorems I had gathered throughout the year and put it together to form a purposeful whole. Paper is broken down into a few sections: Problem defining, Mathematical prerequisites, Construction of the model, Conclusion and use in ML. My goal was to write it so that every high school student with decent math background can understand and follow.
PCA algorithm is an unsupervised learning algorithm used for data dimensionality reduction and it's often applied before the actual supervised learning algorithms such as Artificial Neural Networks. The goal of the PCA is to significantly decrease the number of dimensions while retaining as much information given in the original data as possible. This becomes very important for the efficiency of the actual learning algorithms, especially when we talk about ANN's because of their complex architecture and vast number of calculations.
In a way, this paper also represents the summary of my work and dedication to mathematics over the last four years. Words cannot express how much gratitude I owe to Brilliant and this incredible community for my improvement in mathematics. Discovering Brilliant was for sure a milestone for me and I could've never written this paper without knowledge and deep understanding I gained here.
So, here's the link, I hope you will have fun reading it and maybe learn something new! I want to hear your impressions, critiques and suggestions in the comment section.