The Data Science Cheatsheet 2.0 is a five-page reference document designed to provide quick access to essential machine learning concepts and algorithms. Written in TeX and available as a PDF, it serves as a study aid for exam preparation, interview readiness, and general reference purposes. The resource is grounded in MIT's machine learning courses 6.867 and 15.072, making it academically rigorous while remaining accessible to learners with basic statistics and linear algebra knowledge.
The cheatsheet covers a comprehensive range of machine learning topics spanning an introductory semester-long curriculum. These include foundational supervised learning methods such as linear and logistic regression, decision trees, random forests, support vector machines, and k-nearest neighbors. It also addresses unsupervised learning techniques including clustering and dimension reduction approaches like principal component analysis, linear discriminant analysis, and factor analysis. Beyond these core algorithms, the document extends to more specialized areas including boosting methods, natural language processing, neural networks, recommender systems, reinforcement learning, anomaly detection, time series analysis, and A/B testing methodology.
The repository is classified across multiple data science and machine learning domains, including data analysis, pandas, statistics, numpy, algorithms, big data, statistical analysis, python, jupyter, visualization, SQL, and R programming. This broad categorization reflects the interdisciplinary nature of the content, though the author deliberately chose not to include detailed Python or SQL coverage. The rationale provided is that the cheatsheet prioritizes algorithms, models, and concepts that remain relatively stable across industries and time, whereas programming languages and data structures vary significantly by job function and are better learned through hands-on practice rather than paper-based reference materials.
The document has undergone iterative development, with the author noting that time series analysis and statistics and probability sections were added after the initial release. The repository includes a list of future additions under consideration, such as data imputation, generative adversarial networks, and graph neural networks, indicating ongoing maintenance and expansion plans. The author explicitly welcomes suggestions and improvements from the community.
The cheatsheet is distributed under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License, permitting free sharing in educational contexts such as classes and review sessions. The repository includes visual previews of sample pages to give potential users a sense of the document's layout and content presentation. The author, Aaron Wang, maintains the resource and offers additional services including resume review, application assistance, and technical consulting. The project was inspired by an earlier data science cheatsheet by Maverick, with this version representing a substantially expanded and improved iteration on that foundation.
