I decided to look into some old projects of mine and came up with some interpolation techniques I was exploring for a geo-spatial experiment I was doing early on in my career.
So, essentially I was exploring how do I map out the AQI data collected from various points in Delhi. I wanted to interpolate data between the collected locations. So, I used an approach which used Kriging and allowed me to calculate points in between.
X and Y axis are geo-loc coordinates scaled down by MinMaxScaling.
Z axis(height) is the value collected at that point scaled for plotting.
As you can see, the blue points are the data we collected and green points were the interpolated points.
When we compared this to Weather.com data, we realized that we were within 2-3% of the measured values, which was actually insane.
The inspiration was that, I was scrolling Grapevine and saw someone use Quantum Computers to implement some algorithm. Kinda inspired me to share this hahaha.