A Simple Guide to the Strange Life of the SIMC^2.0 Endeavour Challenge

Sometimes, when we open a website, we find ourselves staring at a completely unknown language. A bit annoying, sure, but nothing a translation app cannot fix. 

Even though this year’s SIMC2.0 Endeavour Challenge, A Linear Algebra Guide to the Strange Life of High-Dimensional Data, is written in English, many readers may still find its concepts challenging at first glance. 

The fifteen-page Challenge revolves around studying higher-dimension databases using linear algebra. Such data often reaches up to thousands of dimensions, yet the most important information is usually concentrated along just a few directions. Take a video for example: millions of pixels with individual colours in every single frame. Completely processing every pixel in each frame would be extremely resource-intensive, yet image compression algorithms preserve the quality by only processing the necessary pixels to form a similar picture. 

The first thing that greets you is a giant wall of text that taunts you for even daring to understand it. The letters swimming before your eyes may as well be hieroglyphics. Its vague nature already dunks the reader into a mild confusion, poking the nerves of the participants before they even begin. 

The three sections – a written report, an oral presentation, and a question-and-answer segment, must all be completed in the span of two days. Though the preamble insists that no prior knowledge of linear algebra is needed, the chunks of information that preview each question introduce the basics of each concept and need time to be completely digested. Hence, the art of critical thinking and problem solving is not lost.  

Firstly, Question One begins with the dynamics between two types of organisms, where a fraction converts to the second type every generation. It’s a simple introduction to linear algebra via observing long-term behaviour of population dynamics using a special type of vectors.  

Question Two follows up on using the Power Method to find the important information within massive amounts of data. It builds on Question One’s concepts, forming an organic progression in difficulty. 

Lastly, the participants need to spearhead a data analysis of a molecular spectroscopy experiment. This lets them apply science and machine learning to spectroscopic datasets. Using the concepts learnt from the other two questions, participants outline the key data like real scientists.   

This year’s Challenge not only tests the participants’ math skills, but also their critical thinking, problem-solving, and computational skills. With the deadline for the written report approaching the very next day, we leave it up to the participants to rack their brains and tackle this Challenge.