Mars Spectrometry 2: Gas Chromatography -- Second place solution

TL;DR

The solution utilized two-dimensional, image-like representations of the competition's chromatography data samples and a number of different Convolutional Neural Network models were trained and ensembled for the final submission.

Abstract

The Mars Spectrometry 2: Gas Chromatography challenge was sponsored by NASA and run on the DrivenData competition platform in 2022. This report describes the solution which achieved the second-best score on the competition's test dataset. The solution utilized two-dimensional, image-like representations of the competition's chromatography data samples. A number of different Convolutional Neural Network models were trained and ensembled for the final submission.

Figures (7)

• Figure 1: Raw intensity values for sample S0801 at m/z=18
• Figure 2: Sample S0801 converted to 2D representation and saved into red channel, where y-axis is mass rows ($0 \le m \le 255$) and x-axis is time columns ($0 \le t \le 191$). The green and blue channels are loaded with $t/191$ and $m/255$ as per Mars1st, correspondingly.
• Figure 3: Same as in Fig. \ref{['fig2']} but divided by maximum column values (mass-normalization).
• Figure 4: Same as in Fig. \ref{['fig2']} but divided by maximum row values (time-normalization).
• Figure 5: Same as in Fig. \ref{['fig2']} but with swapped mass and intensity values, and without the mass and time positional channel encodings. The red color component was proportional to mass, where a mass value of 255 corresponded to the maximum color red value of 255 (in uint8 RGB color encoding).