Skip to content

An analysis of horse colic data using machine learning approach

Notifications You must be signed in to change notification settings

Kok-Herng/horse-colic-data-with-ML

Repository files navigation

horse-colic-data-with-ML

Colic is a condition of abdominal pain in horses and also referred as gastrointestinal disorders. The raw data is obtained from UCI machine learning repository

Aim:

  • To classify was the lesion surgical or non-surgical
  • To perform classification using Artificial Neural Network (ANN), Support Vector Machine (SVM) and Random Forest (RF)
  • To perform clustering using Self-Organizing Map (SOM)

The pre-processing steps carried out were:

  • Remove columns with more than 40% missing values
  • Replace missing values with median value of that particular column for numerical data
  • Replace missing values with nearest non-missing value for categorical data

Conclusion:

  • ANN is the best classifier in this case, achieving 94.84% classification accuracy compared to SVM (88.2%) and RF (87.3%)
  • SOM was able to cluster the data into two distinct group

About

An analysis of horse colic data using machine learning approach

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages