machine learning - Adding one more feature to my feature set which has no effect in calculation and act as a distiguishable feature -
i have of problem follow:
i have 10000 thousand of tweets , and have features labeled 1 or 2 . wanna add feature problem here:
i want give each tweet (a feature)a unique id according user posted tweet, if user posted 3 posts these 3 posts same id since same user posted them way can more sure in classifying these tweets in same group since can claim if of these tweets assigned same label newly coming tweet more probable in same group.also using decision tree , naive bayes question make sense since feature not numeric , has no effect in calculation , acts dummy feature used distinguishing tweets?
Comments
Post a Comment