Posts
-
Fast overlap finding for geographic datasets using GeoPandas
Technical post, 20 minute read
Comparison of a slower, brute-force method of checking for intersections within a geographic dataset, with a faster method that leverages a technology known as a spatial index or R-tree, using the GeoPandas Python package. -
Mapping the global potential of natural reforestation projects using ground observations, remote sensing, and machine learning
Conceptual post, 20 minute read
Creation of a global model for natural forest regeneration rates, with a Google Earth Engine app demonstrated for Africa. -
Supercharging customer touchpoints with uplift modeling
Technical post, 33 minute read
An introduction to a powerful way of predicting individual treatment effects, with synthetic data in Python using pandas and XGBoost. -
Designing an energy arbitrage strategy with linear programming
Technical post, 30 minute read
How to use a classic mathematical optimization technique to guide the operation of a grid-connected battery and maximize profit. -
Visualizing a balance history from transaction data using pandas and Plotly
Technical post, 18 minute read
A practical post on data wrangling and visualization that could be applied in a professional setting, or to your personal finances. -
Why decision trees are more flexible than linear models
Conceptual post, 7 minute read
This blog post will examine a hypothetical dataset of website visits and customer conversion, to illustrate how decision trees are a more flexible mathematical model than linear models such as logistic regression. -
Overfitting, underfitting, and the bias-variance tradeoff
Technical post, 25 minute read
Explore these foundational machine learning concepts and learn how to use regularization to leverage the bias-variance tradeoff to improve model performance, through an illustrative example of polynomial fitting in Python and scikit-learn.
subscribe via RSS