ocsvm-transit-detection

The source code for a One-Class SVM model designed to detect exoplanet transit events.

https://github.com/jakobroche/ocsvm-transit-detection

Science Score: 26.0%

This score indicates how likely this project is to be science-related based on various indicators:

○
CITATION.cff file
✓
codemeta.json file
Found codemeta.json file
✓
.zenodo.json file
Found .zenodo.json file
○
DOI references
○
Academic publication links
○
Academic email domains
○
Institutional organization owner
○
JOSS paper metadata
○
Scientific vocabulary similarity
Low similarity (4.9%) to scientific vocabulary

Last synced: 10 months ago · JSON representation

Repository

The source code for a One-Class SVM model designed to detect exoplanet transit events.

Basic Info

Host: GitHub
Owner: JakobRoche
Language: Python
Default Branch: main
Homepage: https://arxiv.org/abs/2407.00504
Size: 11.7 KB

Statistics

Stars: 0
Watchers: 1
Forks: 0
Open Issues: 0
Releases: 0

Created almost 2 years ago · Last pushed 11 months ago

Metadata Files

Readme Codemeta

README.md

modelcondensed.py is the Python file that compiles the One-Class SVM model for transit detection.

To compile the model: Open modelcondensed.py in a Python editor. Download the KOIQ16long dataset file. This houses the PDCSAP flux files that the model uses as a training dataset. Note that cloning the repository to access the dataset file may not work. To ensure everything runs smoothly, manually downloading the file and referencing it from there is preferable.

Change the directorypath value in the code to wherever the KOIQ16_long dataset file is stored.

Once the code is run, it should preprocess the data, train and evaluate the model.

Owner

Login: JakobRoche
Kind: user

Repositories: 1
Profile: https://github.com/JakobRoche

CodeMeta (codemeta.json)

{
  "@context": "https://doi.org/10.5063/schema/codemeta-2.0",
  "@type": "SoftwareSourceCode",
  "name": "OCSVM-Transit-Detection: One-Class SVM model for exoplanet transit detection",
  "description": "This One-Class Support Vector Machine (SVM) model detects exoplanet transit events. One-class SVMs fit data and make predictions faster than simple CNNs, and do not require specialized equipment such as Graphics Processing Units (GPU). The code uses a Gaussian kernel to compute a nonlinear decision boundary. After training, OCSVM-Transit-Detection requires that lightcurves classified as containing a transit have features very similar to the lightcurves in the training dataset, thus limiting misclassifications.",
  "identifier": "ascl:2506.013",
  "author": [
    {
      "@type": "Person",
      "givenName": "Jakob",
      "familyName": "Roche",
      "email": "jakoblaszloroche@usf.edu"
    }
  ],
  "citation": "https://ui.adsabs.harvard.edu/abs/2025ascl.soft06013R",
  "codeRepository": [
    "https://github.com/JakobRoche/OCSVM-Transit-Detection"
  ],
  "version": "2.0",
  "license": "https://spdx.org/licenses/MIT.html"
}

ecosyste.ms

Data

Tools

Indexes

Applications

Experiments

Open Source Science