PinnedJustin SwansburginBetter ProgrammingTwo Ways to Apply OpenAI’s GPT and Other LLMs to the Non-Text Features in Your ML ModelsLeverage prompt engineering and open-source embeddings to enrich your datasets and inject more domain context into your modelsApr 16, 20232Apr 16, 20232
PinnedJustin SwansburginGoPenAIHow to Uncover Feature Interactions in your Machine Learning Models with Two-Way Partial Dependence…A comprehensive walkthrough on how to discover, visualize, and interpret feature interactions in your ML modelsApr 12, 2023Apr 12, 2023
PinnedJustin SwansburgVisualizing Shapley Values Over TimeThe ML insight you don’t know aboutApr 2, 20231Apr 2, 20231
PinnedJustin SwansburginBetter ProgrammingHow ChatGPT Can Improve Your ML ModelsLeverage generative AI to improve your model performance with this simple trickApr 18, 20234Apr 18, 20234
PinnedJustin SwansburgA New Way To Visualize Lift Charts with Shapely ValuesInventive new technique to combine Shapely value explanations with your lift charts to enhance model explainabilityMar 26, 2023Mar 26, 2023
Justin SwansburgNo Free Lunch: The Importance of always Testing many Machine Learning ModelsAn in-depth look into which machine learning models perform the best on real-world problemsJun 18, 20231Jun 18, 20231
Justin SwansburgSupervised Anomaly Detection: A Better Way to Model AnomaliesStandard anomaly detection models are hard to evaluate and often fail to reliably catch anomalies. Try this new supervised approach that…Jun 5, 20232Jun 5, 20232
Justin SwansburgThe Brave New World of Feature Engineering: How LLMs Can Enrich Your DataThree fun tricks on how to engineer new features with LLMs that you can include in your modeling pipelinesMay 17, 20231May 17, 20231
Justin SwansburgHow to Use LLMs to Build Better Clustering ModelsA quick guide on a new way to leverage large language models to improve the separation across your clustersMay 13, 20232May 13, 20232
Justin SwansburginGoPenAIThe Hidden Risks of Leveraging OpenAI’s GPT in your Machine Learning PipelinesHow to address the issues of randomness, target leakage, hallucination, and topic drift when applying GPT and other LLMsApr 23, 20232Apr 23, 20232
