0
$\begingroup$

Let us assume that we have a certain number of features that are weighted with some parameters . The features could be the different skills that belongs to a job candidate applying for a job vacancies . The decision to give a job to the single candidate is based on a weighted sum of a different features .So we have a neural network model on which we have to learn the matrix that associate the variable x = features that a candidate has to y = ( hire/not hire) I was wondering what is the name in literature of this problem and where i can find suggestion to solve it. Many thanks

$\endgroup$
0
$\begingroup$

its a simple DNN problem when you have x features and y labels you can have some data for NN training and then NN can be used to classify other candidates. I have a program already made for this plz comment if you need one .

| improve this answer | |
$\endgroup$
  • $\begingroup$ You are completely right but more than the algorithm I was wondering if exists a dataset for the training of the algorithm. I tried to google it but I think it is difficult to find a company dataset that has been modelled like a DNN problem like this. This is main reason that pushed me to ask how others tackled the same problem. $\endgroup$ – AntonYellow Jul 7 '19 at 20:43
  • $\begingroup$ Hi anton u r also right about the dataset as its kind of proprietary to companies but may be its an opportunity to create one. And than we can open source it as well. I know how to extract features from text and if you have enough CVs than we can go for this. $\endgroup$ – P K Jul 8 '19 at 5:33
  • 1
    $\begingroup$ Interesting proposal , my idea is that we need two abstractions to tackle the problem . The first is the features extraction from resume in a smart way and the second is the creation of the complex features from the simple feature extracted from the resume. Then we can go unsupervised I Suppose. $\endgroup$ – AntonYellow Jul 8 '19 at 16:08
  • $\begingroup$ the second complex feature creation from simple features is also possible and there are quite a number of research paper on it. So how should we go about it now $\endgroup$ – P K Jul 8 '19 at 16:23
  • $\begingroup$ Ok then after the creation of a vector containing N predictors like ( age, salary requested , number of job transition in the past 5 years etc extracted from the resume ) we have N weights that measure the importance of the predictors for the different companies . The final score is one row of the matrix . Is this feasible in your opinion to populate a matrix like this ? $\endgroup$ – AntonYellow Jul 13 '19 at 9:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.