Machine Learning Articles

Kurtis Pykes

Updated on May 22, 2025

Cohen’s kappa is a statistical metric that measures the reliability of two raters who are evaluating the same thing, accounting for the possibility that they could agree by chance. Here’s how it works and how to calculate it.

Akash Desarda

Updated on May 22, 2025

AdaBoost (Adaptive Boosting) is a machine learning boosting technique used as an ensemble method to adjust the weights of training samples and combine multiple weak classifiers into a single strong classifier.

Kamil Krzyk

Updated on May 22, 2025

A cost function in machine learning quantifies the error between a model’s predicted values and actual values, letting us measure how well a model's parameters fit given data. I’ll introduce you to two often-used regression cost functions: MAE and MSE.

Francesco Bellelli

Updated on May 22, 2025

A Voronoi diagram (or Dirichlet tessellation or Thiessen polygons) is a type of tessellation pattern where points on a plane are divided into exactly n number of cells, which enclose a region of the plane that is closest to each point.

Sara A. Metwalli

Updated on May 22, 2025

Data validation refers to verifying the quality and accuracy of data before using it. These are the main types of data validation, the pros and cons of the process and tips for how to perform data validation.

Peter Grant

Updated on May 22, 2025

The bias-variance tradeoff describes the inverse relationship between bias and variance, where increasing one decreases the other. Here’s how to strike a balance between the two, so a model learns enough details about a data set without picking up noise.

Perez Ogayo

Updated on May 22, 2025

A CUDA Error: Device-Side Assert Triggered can either be caused by an inconsistency between the number of labels and output units in a model or an incorrect input for a loss function. Follow this guide to fix it. 

Alexandre Zajic

Updated on May 22, 2025

Akaike Information Criterion (AIC) is a metric with a single number score that measures which machine learning model is best for a given data set, in comparison to other models for the same set. Here’s what you need to know.

Richmond Alake

Updated on May 22, 2025

Cosine similarity measures the similarity between two non-zero vectors by calculating the cosine of the angle between them. Here's the basics behind cosine similarity and how it is used across different areas of machine learning.

Vegard Flovik

Updated on May 22, 2025

Graph theory is the study of graph data structures, which model object relationships using vertices connected by edges. It is a helpful tool to quantify and simplify complex systems.