Published inTowards Data ScienceThe Key to Smarter Models: Tracking Feature HistoriesCapture context and improve predictions with historical dataDec 31, 20245Dec 31, 20245
Published inTowards Data ScienceA Simple Strategy to Improve LLM Query GenerationDynamic Few-Shot PromptingAug 6, 20243Aug 6, 20243
Published inTowards Data ScienceEarly Stopping: Why Did Your Machine Learning Model Stop Training?Why most models are small and LLMs are largeMay 11, 20242May 11, 20242
Published inTowards Data ScienceCausal Machine Learning: What Can We Accomplish with a Single Theorem?Exploring and exploiting the seemingly innocent theorem behind Double Machine LearningMar 30, 20242Mar 30, 20242
Published inTowards Data ScienceCategorical Features: What’s Wrong With Label Encoding?Why we can’t arbitrarily encode categorical featuresNov 20, 20238Nov 20, 20238
Published inTowards Data ScienceBuilding PCA from the Ground UpSupercharge your understanding of Principal Component Analysis with a step-by-step derivationAug 7, 20231Aug 7, 20231
Published inTowards Data ScienceAliasing: Your Time Series is Lying to YouAn intuitive introduction to signal aliasing with PythonJul 12, 20231Jul 12, 20231
Published inTowards Data ScienceA Performant Way to Query Streaming DataImproving a real-time streaming project with ksqlDBMay 22, 20231May 22, 20231
Published inTowards Data ScienceLearning Math Through Code: DerivativesGain a deeper understanding of derivatives with PythonMar 20, 2023Mar 20, 2023
Published inTowards Data ScienceAnother (Conformal) Way to Predict Probability DistributionsConformal multi-quantile regression with CatboostMar 8, 20235Mar 8, 20235