Closed rbutler97 closed 6 years ago
Tried linear regression, but because the data forms approximately a bell curve every year, when we ran the algorithm it gave a line with no slope that returned approximately the average UV index for every data point.
Made extensive attempts with Gaussian process classification, which produced inaccurate results and would not work with a data set of over a year, as well as trying logistic regression, neural networks, and isotonic regression, none of which produced accurate results. We used the scikit library for all of these algorithms.
Possible methods: Linear Regression, Logistic Regression, Polynomial Regression, Neural Networks (Use SciKit)