Hi Kumar,

You should choose classifier which copes with nominal features, such as decision tree or random forest. Most classifiers require numerical features (neural network, SVM, Bayes).

Best regards

Hi Kumar,

You should choose classifier which copes with nominal features, such as decision tree or random forest. Most classifiers require numerical features (neural network, SVM, Bayes).

Best regards

Hi 

There are various ways to do this:

First, you can convert the feature values like Java, Node.js, MVC etc to 1,2 3,4 etc and declare the variable as nominal or categorical.

Second, You can convert the categorical feature as binary variable like 001, 011, etc. Then it can be given as input to the algorithm.

I hope I have clarified the question. In case of any doubts, please feel free to contact me.

Thank you.

Best regards,

Dr.Indrajit

You could also use multinomial model approach, please have a look to my article 

Logistic Regression Models. In page 10 you'd  see an example of multinomial model using R.

Assign numerical values to the skills. The outputs are important for your training as well. If the output is may be that an applicant is selected or not, then assign binary values like 1 for selected and 0 for not selected. Cross-validate the performance of two or more classifiers on your dataset and choose the one that have the best out of sample error.

Any classifier that admits features whose values are not numerical can be used in principle. For example a decision tree learner.

Use ranking based on each class..like Engineering ( E) ::: then E1, E2, E3...like that and continued.....

 I suggest the second option given by Dr. Indrajit Mandal (the binary representation). I also suggest to use features selection on this binary representation. Another option is to use real values between zero and one for each skill to give an importance to the degree of the skill.

Hi Kumar,

you can conver the "skills" feature into ranks say [1..5],

where 1: for poor or non-skills 2: Low-skills & experience 3: Moderate 4: High-levels of skills, and 5: very high-level of skills & experiences

Regards

Thank you Dr. Indrajit Mandal for your response. I would appreciate if you could elaborate the solution or else provide any pointers for that.