Predicting movie ratings

Ever wondered why a movie got rated so highly when you thought that it was terrible? Well in this Kaggle kernel I explore IMDB movie ratings and the different variables that contribute to a movie’s success. I conclude that director and cast are the most important things to consider, and I develop a model to predict a movie’s ratings.

 

Check it out here for details: Kaggle

Assessing global happiness

What makes you happy? Well there is a thing called the world happiness index which is an attempt at quantifying how happy citizens of world countries are. In this Kaggle kernel, I relate happiness to a bunch of environmental variables, playing with the idea that people in sunny countries are likely to be happier than those in cloudy/rainy ones.

Check it out here for details: Kaggle

Fractional cover

South Africa: rainbow ground cover. A Landsat satellite pixel covers a 30 x 30m area on the ground. The reflectance from this square encompasses a range of substances – e.g. bare ground, water, green grass, concrete etc. Spectral unmixing allows one to decompose a measured spectrum of light into sub-specta which represent different ground surfaces. Here I have differentiated the fractional contribution of bare ground (red), green vegetation (green) and brown vegetation (blue) to individual Landsat pixels over South Africa. The resulting image gives an intuitive indication of ground cover.

For other examples check out this Instagram page.