Skip to main content

Timeline for ML - Text input data

Current License: CC BY-SA 4.0

9 events
when toggle format what by license comment
Jan 15, 2019 at 2:22 comment added Carl Suggest that you are using a decision tree of the if...then type currently. Either one figures out what decision tree is being used or one uses a neural net to attempt to duplicate it, without thinking it through. Personally, I would use the former method as it is more nuts and bolts and can be altered without having to train a network all over again.
Jan 14, 2019 at 23:34 comment added ranit.b @Carl - You can think of service group as specific application areas or division within a category. Ex. For category 'Application', we can have multiple service group like - Microsoft applications, or Oracle database, or AWS, or GitHub, or etc.
Jan 14, 2019 at 20:48 comment added Carl It is unclear to me at least what service group is. It is unclear to me how many teams there are and what they do. It is unclear how these variables are distributed. It is unclear what causes an assignment to which team, and whether or not the assignments are currently being performed correctly. It is unclear what type of statistical model (linear, nonlinear, type) can be used to predict assignments.
Jan 13, 2019 at 10:45 comment added ranit.b @Bjorn - Not yet. I had been focusing more into regression and classification mechanisms. But, why do you suggest NLP ? Please advise.
Jan 13, 2019 at 10:43 comment added ranit.b @Carl - Not 'service group'. I want to predict the 'Team' where the ticket has to be assigned. I have historical data where based on the above features, service desk team had assigned (manually) the tickets to various teams.
Jan 13, 2019 at 8:20 comment added Björn Have you had a look at the natural language processing (NLP) literature or e.g. the fast.ai course?
Jan 13, 2019 at 7:51 comment added Carl Is the variable service group the eventual assignment that you want to predict?
Jan 13, 2019 at 1:40 review First posts
Jan 13, 2019 at 7:51
Jan 13, 2019 at 1:35 history asked ranit.b CC BY-SA 4.0