Projects Here's two of my Python projects. Climate Forecasting is a project which uses time series analysis to predict the next 500 years' temperature profile based on 800,000 years of historical data taken from ice core samples in the Antartic. Interestingly, it predicts we 'should' be in the midst of a cooling trend. The project compares various RNN and LSTM models to varieties of linear regression models. Interestingly, a linear regression model does the best job. There is also a project which attempts to forecast hurricane trajectories two days in advance. It uses NOAA hurricane data over several decades to once again train both linear regression models of varying complexity and varieties of LSTM and RNN neural network models. The linear model outperforms the neural network model once again, probably because it more quickly learns the Coriolis force based on the somewhat limited data. Both of these projects run in jupyter notebook. XtraMLTools is some specialized machine learning classes I wrote. They extend the functionality of sklearn's regression and logistic regression classes. First they make it easy to perform polynomial regression on user-specified columns of data. And second, they allow one to do a separate regression, basically, for each value of user-specified columns of categorical data (possibly multiple columns). In effect, it is combining decision trees on user-specified categorical data with polynomial regression on numerical data.