Multi-Class and Multi-Label Classification using Support Vector Machines
Multi-Class and Multi-Label Classification Using Support Vector Machines on Anuran Cells Data Set. The dataset contains three labels and each label has multiple classes.
- Trained SVM for each labels using Gaussian Kernels and one versus all classifiers. Also, trained the model using L1-penalized SVMs.
- Used 10 fold CV to determine the weight of SVM penalty and width of the Gaussian Kernel.
- Used SMOTE to remedy class imbalance and trained the classifiers.
- Ski-kit learn
- Pandas Library
- Python
Dataset from (https://archive.ics.uci.edu/ml/datasets/Anuran+Calls+%28MFCCs)