Start here in Data Science
The most useful guides in this topic, not just the most recent.
Data Science Guides & Tutorials
Roadmaps, salaries, projects and interview prep for aspiring data scientists — from statistics and machine learning to model deployment.
Data Science articles
Ensemble Learning: Bagging, Boosting and Stacking Explained
Ensemble learning explained with measurements on one real dataset: bagging steadies an unstable decision tree, boosting turns 89%-accurate stumps into a 97% model, and stacking combines model types, yet plain logistic regression still edges ahead. What each fixes, why diverse errors matter, and how to choose.
How to Build a Data Science Portfolio That Gets Interviews
How to build a data science portfolio designed for the few minutes a reviewer spends on it: which three projects to show, a README template that puts the result first, repository habits that signal professional work, when a live demo is worth it, and the red flags that get portfolios skipped.
Hyperparameter Tuning: Grid Search vs Random Search vs Bayesian
Grid search, random search and Bayesian optimisation compared on one real model with the same budget of 24 candidates: how close they came, why the differences sat inside the noise between folds, why random search explores more, and a second model where tuning made the difference between 61% and 93% accuracy.
Time Series Forecasting: A Beginner-Friendly Introduction
A beginner-friendly introduction to time series forecasting on 36 years of real monthly CO2 measurements: what trend and seasonality are, why simple benchmark methods matter, how Holt-Winters combines both patterns, and how to test forecasts honestly by repeatedly forecasting the next year from past data only.
How to Read a Machine Learning Research Paper
How to read a machine learning research paper without drowning in it: the three-pass method applied to the Transformer paper, the questions to ask of every results table, the difference between what a paper shows and what it suspects, and a small NumPy re-implementation that checks one of its claims.
Computer Vision Basics: From Pixels to Predictions
Computer vision from the ground up: what an image is to a computer, how a convolution filter finds edges, a small convolutional network trained on handwritten digits against a pixel baseline, the main vision tasks from classification to segmentation, and why transfer learning is the default.