Explores the techniques for data visualization in machine learning using Elixir, utilizing libraries such as VegaLite and Axon. It includes methods for setting up dependencies, creating scatter plots, facetted scatter plots, and multiple scatter plots with the Iris dataset. Additionally, it demonstrates how to track and plot training metrics like loss and accuracy, both post-training and in real-time. The post concludes with a look into Axon's native plotting capabilities and hints at further exploration in Elixir's ML ecosystem.
Table of contents
Project Setup #Basic Scatter Plot #Facetted Scatter Plot #Multiple Scatter Plot #Training Output Data #Collecting the Loss History #Plotting the Loss and Accuracy #Plotting the Training data in Real-Time #Axon’s Native Plotter #Closing Thoughts #Sort: