Preprocessing:

• Cleaned the data-set and used feature selection to choose best features.
• Performed Normalization

Development:

• Developed a multi-label classification system based on learning from multi-class data.

Algorithms Used:

• Tried several approaches such as Logistic Regression, SVM, Random Forests (with parameter optimization), out of which Random Forests gave the best results.

Conclusion:

• Successfully distinguished between product categories.

Team:

• Rohit Nair
• Sohil Jain