Subscribe to DSC Newsletter

Suggestions for handling a multi-label multi-class unsupervised classification recommendations

Hi, I'm handling a  problem as shown in the below diagram, I need to predict the required output from the user inputs as shown below. I would like to know some suggestions regarding the best Machine learning approach to be used for solving this task.

Dataset for learning:

User Input:

Required Output:

About Dataset:

1. Every column includes different levels of value.

2. All are categorical values.

3. This is an unsupervised dataset.(There is no fixed y variable)


1. User may change the order of giving input variables.

2. User may change the number of input columns 

    Eg. input= column1,column2    =>   output= column3, column4, column 5

           input= column1,column5, column3  =>  output= column2, column4

NOTE: We have tried the following algorithms, but looking forward for a better approach to handle this task. 

1. Aprori , Eclat, FP-Growth Algorithms

2. Multi-class multi-label classification: Binary relevance, classifier chains with Random Forest and    Decision Tree

Tags: Deep Learning, Machine Learning, NLP, Recommendations

Views: 646

Reply to This


  • Add Videos
  • View All

© 2020   Data Science Central ®   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service