Logo J. Marcus Hughes
    Logo Inverted Logo
    • Posts
    • Welcome
    • Anomaly Detection
      • Atypicality
      • Axis-aligned RRCF artifacts
      • goal of nonstationary
      • Mostly noise training
    • Assorted Coding
      • flood
      • inclusion check
      • mood tracking
      • np.sum speeds in Cython
      • sliding puzzles
      • truncating numbers
      • Ulam Warburton
    • Assorted ML
      • fake APOD generation
      • TSS vs f1
    • Denoising
      • denoising autoencoder
    • Paper Reading
      • State of Astro-informatics
    • Photography
      • New Instagram
    • Solar
      • AIA temporal coloring
      • automatic solar YouTube
      • EUI artifact
      • HEK thematic maps
      • SUVI calendar heatmap
      • Thematic Map Paper
    • Teaching
      • peer review activity
    Hero Image
    Axis aligned artifacts

    There are artifacts created by choosing axis aligned cuts in robust random cut forests, similar to what was noted with IsoForest.. Left: Original data distribution. Right: Learned co-displacement, darker is lower. Notice the echoes around (10,-10) and (-10, 10) If instead of either of these, you use the depth in the robust random cut forest, you get what’s shown above. The first two examples are recreated by the code below:

    October 7, 2019 Read
    Hero Image
    Training an autoencoder with mostly noise

    I am working on a project where we wish to use anomaly detection to find what image patches have structure and which don’t. As an aside, I ran an experiment on MNIST. You have 500 images of fives. You have 5000 images that are pure noise. You train a deep convolutional autoencoder. What you end up with is the following reconstruction: The top row are the inputs and the bottom row are the reconstructions.

    September 17, 2019 Read
    Hero Image
    Goal of Anomaly Detection in Non-stationary Data

    I was explaining anomaly detection in non-stationary data to someone and threw together this crude example figure. The blue points are nominal and represent 90% of the points. The red are anomalous and represent 10% of the points. In this example, the red data is stationary while the blue passes through it. Thus, it would be very difficult to differentiate the red and blue points when they overlap. However, even if we only had a few frames of this video, we would like to be able to realize there are two dynamics going on.

    September 9, 2019 Read
    Hero Image
    Atypicality Presentation Recap

    Yesterday, I gave a presentation introducing the ideas of atypicality to the Monteleoni research group. These are the slides and handwritten notes. I plan to explore this idea further and write up better LaTeX notes, which I will then share as well. For now, the idea of atypicality centers around using two coders: one trained to perform best on typical data and one that is universal and not data specific. A sequence is atypical if its code length using the typical coder is longer than the universal coder, i.

    May 3, 2019 Read
    Navigation
    • About
    Contact me:
    • hughes.jmb@gmail.com

    Toha Theme Logo Toha
    © 2022 Copyright James Marcus Hughes
    Powered by Hugo Logo