Matthew Slocum

Background

About Matthew Slocum

I began my career as an ecologist. I joined graduate school, did my PhD, and discovered that my biggest passion—the thing that drives me day to day—is data. And by data, I mean all aspects of data: how it is massaged and analyzed, how it is plotted and communicated, how it is viewed philosophically, and how all these different aspects feed into each other to create a useful model. By starting with ecological data, I learned to handle complex, multivariate, and messy data sets. In analyzing these data, I have become skilled at plunging into data pools, wrestling with the critters inside, and pulling out the most important species. I know how to develop tools to create models that are simple and easily communicated, yet not so simple as to be unhelpful. I can develop these tools because I have become familiar with many techniques, thereby allowing me to attack problems from multiple directions. I have learned to model as directed by theory; I don’t let the method do the directing. In the 10+ years I have been working in the oil and gas industry, I have adjusted my skill set so I can handle much larger data sets, quick changes in scope, and different personalities and viewpoints. A list of my skills include: • Merging and cleaning data from diverse sources into a coherent foundation using Python. • Classic statistics: basic and advanced regression; ANOVA; cluster analysis and other ordination techniques; PCA; structural equation modeling; modeling selection. • Spatial modeling: Kriging • Machine learning: decision trees; random forests; neural nets; reinforcement learning; transformers. • Understanding the theory behind modeling to avoid common errors (e.g, overfitting) using cross-validation and other strategies. • Listening to clients to understand their main questions, and then translating these to solutions. • Communicating in reports and presentations using clean, attractive graphics and language.