AbstractPattern recognition and machine learning are becoming integralparts of algorithms in a wide range of applications. Di erent algo-rithms and approaches for machine learning include di erent tradeo sbetween performance and computation, so during algorithm develop-ment it is often necessary to explore a variety of di erent approachesto a given task. A toolbox with a uni ed framework across multiplepattern recognition techniques enables algorithm developers the abilityto rapidly evaluate di erent choices prior to deployment. MATLAB isa widely used environment for algorithm development and prototyp-ing, and although several MATLAB toolboxes for pattern recognitionare currently available these are either incomplete, expensive, or re-strictively licensed. In this work we describe a MATLAB toolbox forpattern recognition and machine learning known as the PRT (PatternRecognition Toolbox), licensed under the permissive MIT license. ThePRT includes many popular techniques for data preprocessing, super-vised learning, clustering, regression and feature selection, as well asa methodology for combining these components using a simple, uni-form syntax. The resulting algorithms can be evaluated using cross-validation and a variety of scoring metrics to ensure robust performancewhen the algorithm is deployed. This paper presents an overview ofthe PRT as well as an example of usage on Fisher’s Iris dataset.