Machine learning research at the intersection of manufacturing and data sciencehttps://www.tvhahn.com/en-ushttps://www.tvhahn.com/posts/building-a-claude-code-skill/https://www.tvhahn.com/posts/building-a-claude-code-skill/How I built a Claude Code skill for publication-quality matplotlib charts over 50 iterations — and what happened when I automated its improvement.Fri, 03 Apr 2026 00:00:00 GMTClaude CodeSkillsMatplotlibData VisualizationAutomationTim von Hahnhttps://www.tvhahn.com/posts/pyphm/https://www.tvhahn.com/posts/pyphm/Why build an open-source package for machinery data?Thu, 10 Mar 2022 14:14:56 GMTCondition MonitoringData ScienceTim von Hahnhttps://www.tvhahn.com/posts/finding-inspiration-in-random-ways/https://www.tvhahn.com/posts/finding-inspiration-in-random-ways/A wellspring of ideasMon, 17 Jan 2022 23:40:00 GMTIdea GenerationInspirationCreativityTim von Hahnhttps://www.tvhahn.com/posts/beautiful-plots-violin/https://www.tvhahn.com/posts/beautiful-plots-violin/Music to my ears... err... eyes? The violin plot is a worthy tool for any data visualization tool box. Let's build one, in Plotly, as we explore historic birth trends in the USA!Tue, 28 Sep 2021 00:00:00 GMTPlotlyViolin PlotData VisualizationBeautiful PlotsCDC Birth DataTim von Hahnhttps://www.tvhahn.com/posts/jupyter-hpc/https://www.tvhahn.com/posts/jupyter-hpc/Using Jupyter Notebooks in a high performance computing environment is easy! This tutorial will show you how.Sun, 20 Jun 2021 19:14:56 GMTHPCJupyter NotebookTutorialData ScienceTim von Hahnhttps://www.tvhahn.com/posts/anomaly-results/https://www.tvhahn.com/posts/anomaly-results/We've trained the variational autoencoders, and in this post, we see how the models perform in anomaly detection. We check both the input and latent space anomaly detection effectiveness.Mon, 31 May 2021 13:29:01 GMTPrecision Recall CurveROC CurveMachine LearningCondition MonitoringVariational AutoencoderTensorFlowAnomaly DetectionMillingTim von Hahnhttps://www.tvhahn.com/posts/beautiful-plots-lollipop/https://www.tvhahn.com/posts/beautiful-plots-lollipop/The lollipop chart is great at visualizing differences in variables along a single axis. In this post, we create an elegant lollipop chart, in Matplotlib, to show the differences in model performance.Fri, 14 May 2021 01:20:01 GMTMatplotlibLollipop ChartData VisualizationCleveland PlotBeautiful PlotsDot PlotTim von Hahnhttps://www.tvhahn.com/posts/beautiful-plots-decision-boundary/https://www.tvhahn.com/posts/beautiful-plots-decision-boundary/How can we communicate complex concepts using data visualization tools? In this first post -- in a series titled 'Beautiful Plots' -- we build an elegant chart demonstrating the decision boundary from a KNN classifier.Sat, 24 Apr 2021 02:05:00 GMTMatplotlibKNNData VisualizationDecision BoundaryBeautiful PlotsTim von Hahnhttps://www.tvhahn.com/posts/building-vae/https://www.tvhahn.com/posts/building-vae/In this post, we'll explore the variational autoencoder (VAE) and see how we can build one for use on the UC Berkeley milling data set. A variational autoencoder is more expressive than a regular autoencoder, and this feature can be exploited for anomaly detection.Fri, 12 Mar 2021 00:00:00 GMTMachine LearningCondition MonitoringVariational AutoencoderTensorFlowAnomaly DetectionMillingTim von Hahnhttps://www.tvhahn.com/posts/milling/https://www.tvhahn.com/posts/milling/Data exploration is an important first step in any new data science problem. In this post we introduce a metal machining data set that we'll use to test anomaly detection methods. We'll explore the data set, see how it is structured, and do some data visualization.Mon, 30 Nov 2020 00:05:01 GMTAIMachine LearningCondition MonitoringData ExplorationData VisualizationSeabornMatplotlibTim von Hahnhttps://www.tvhahn.com/posts/anomaly-detection-with-autoencoders/https://www.tvhahn.com/posts/anomaly-detection-with-autoencoders/Machines fail. Anomaly detection, using an autoencoder, can be used to identify failures that have never been seen before.Tue, 10 Nov 2020 00:00:00 GMTAIMachine LearningCondition MonitoringAutoencodersAnomaly DetectionTim von Hahnhttps://www.tvhahn.com/posts/feature-engineering-examples/https://www.tvhahn.com/posts/feature-engineering-examples/Feature engineering can be powerful tool, but it doesn't have to be complicated. Often, the simple solution is best.Sun, 11 Oct 2020 19:14:56 GMTMachine LearningCondition MonitoringFeature EngineeringTotal Harmonic DistortionSignal ProcessingLogistic RegressionMining IndustryMobile EquipmentTim von Hahnhttps://www.tvhahn.com/posts/data-driven-methods/https://www.tvhahn.com/posts/data-driven-methods/In part two, we give an overview of the two data-driven approaches in condition monitoring; that is, feature engineering and end-to-end deep learning. Which approach should you use? Well, it all depends...Wed, 07 Oct 2020 00:00:00 GMTManufacturingMachine LearningCondition MonitoringFeature EngineeringEnd-to-End Deep LearningDeep LearningTool WearTim von Hahnhttps://www.tvhahn.com/posts/business-case/https://www.tvhahn.com/posts/business-case/Why care about data-driven condition monitoring techniques? In this post, we articulate the business case for using these advanced methods in an industrial environment.Tue, 29 Sep 2020 13:29:01 GMTManufacturingMachine LearningCondition MonitoringTim von Hahnhttps://www.tvhahn.com/posts/what-is-school-for/https://www.tvhahn.com/posts/what-is-school-for/What is school for?Thu, 07 Feb 2019 19:14:56 GMTPersonalEducation ReformPassionTim von Hahn