Resampling methods are a crucial tool used commonly in modern statistics and data science. These methods involve taking repeated samples from a training dataset and refitting a model of interest on each individual sample to obtain additional information about the fitted model. These methods allow us to learn new information... [Read More]

In this post, I’ll be exploring quadratic discriminant analysis. I’ll compare and contrast this method with linear discriminant analysis, and work through an example using scikit-learn and the slimmed down Titanic dataset from one of my prior posts on logistic regression. [Read More]

This post is the second in a series on linear discriminant analysis (LDA) for classification. In the first post, I introduced much of the theory behind linear discriminant analysis. In this post, I’ll explore the method using scikit-learn. I’ll also discuss classification metrics such as precision and recall, and compare... [Read More]

This post is the first in a series on the linear discriminant analysis method. In this series, I’ll discuss the underlying theory of linear discriminant analysis, as well as applications in Python. [Read More]

I recently read through this excellent Medium article about the ColumnTransformer estimator in scikit-learn and how it can be used in tandem with Pipelines and the OneHotEncoder estimator. To strengthen my own understanding of the concept, I decided to follow the post with my own working example, and summarize the... [Read More]

This post is the second in a series on the logistic regression model. In this post, I’ll work through an example using the well known Titanic dataset, scikit-learn and statsmodels. I’ll discuss one-hot encoding, create a 3D logistic regression plot using Plotly, and demonstrate multiclass logistic regression with scikit-learn. [Read More]

This is the first post in a series on the logistic regression model. The structure of this post was influenced by the fourth chapter of An Introduction to Statistical Learning: with Applications in R by Gareth James, Daniela Witten, Trevor Hastie, and Robert Tibshirani. [Read More]

This is the fifth post in a series on the multiple linear regression model. In previous posts, I introduced the theory behind the model, exploring using Python’s scikit-learn and statsmodels libraries, and discussed potential problems with the model, such as collinearity and correlation of the error terms. [Read More]

This post is the fourth in a series on the multiple linear regression model. In previous posts, I explored the topic, including methods of relaxing various assumptions made by the model. I also performed a comparison of Python’s scikit-learn and statsmodels libraries for multiple linear regression. [Read More]

This post is the third in a series on the multiple linear regression model. In previous posts, I introduced the multiple linear regression model and explored a comparison of Python’s scikit-learn and statsmodels libraries. However, both of these previous posts exclusively explored quantitative predictors. [Read More]