Detecting Pulsar

Introduction to Data Science - Udacity Project

Dataset information and files can be downloaded from: https://archive.ics.uci.edu/ml/datasets/HTRU2

Medium blog post explaining the steps: https://medium.com/@hariss096/detecting-pulsar-26374beda790

Libraries used:

• pandas
• sci-kit learn
• scipy
• numpy
• matplotlib

Motivation As a machine learning student who watches documentaries related to cosmology and astrophysics in his spare time, this dataset is perfect for me to learn and have fun along the way.

Files inside the repository

• pulsar_stars.csv (Downloaded from UCI Machine learning repository, linked provided above).
• code.ipynb (A jupyter notebook file that follows a CRISP-DM approach for analyzing the dataset).
• .gitignore (To exclude ipynb checkpoints folder from being pushed into remote).

Insights The modelling helps in identifying if the detected object is a pulsar or not (with 95% precision) when following 8 values of the object are provided:

• Mean of integrated profile
• Standard deviation of integrated profile
• Excess kurtosis of integrated profile
• Skewness of integrated profile
• Mean of DM-SNR curve
• Standard deviation of DM-SNR curve
• Excess kurtosis of DM-SNR curve
• Skewness of DM-SNR curve

Details about each of these features can be found in this book: http://www.scienceguyrob.com/wp-content/uploads/2016/12/WhyArePulsarsHardToFind_Lyon_2016.pdf The formulas for each of the features are listed in page number 222 of this book.