0

I just chosen a data set with the following attributes.

  • Zipcode (1000, 4000) only 2 zip codes I have chosen.
  • City (Sydney, Brisbane) only 2 cities I have chosen.
  • Optus
  • Telstra

Just to represent in a area how many people are using optus network and how many people are using Telstra network.

I have 100 records and I calculated Mean, Std. Dev and the count.

Now, stuck what all should I find for NavieBayes apart from mean & std deviation? and if possible can anyone provide with a sample formula too. I am working on C#.

And, calculating precision required?

Thanks.

Kerry
  • 35
  • 2
  • 9
  • http://stackoverflow.com/questions/15132205/implementing-naviebayes-in-c-sharp ? (Both with "Navie"... shouldn't it be "Naive" in the sense of "simple"/"blue eyed"? Oh... and both by you. Hm.) – Corak Mar 04 '13 at 11:06
  • @Corak - I never worked on NavieBayes before. I am confused what all calculations are required in order to perform this classification. I was googling for past one week but couldn't find proper answer – Kerry Mar 04 '13 at 11:10
  • 1
    Maybe searching for "naive" instead of "navie" will yield better results. Also: have you tried [Accord](https://code.google.com/p/accord/)? As I understand it, it has a working implementation that you can learn from. – Corak Mar 04 '13 at 11:17
  • I am not allowed to use any existing libraries. I should code on my own. I am just looking for what all steps to be implemented to call the problem as Naive Bayes Classification. I don't have any knowledge on it but unfortunately its my assignment. I have enough coding skills to code the complete program on my own but just lack of knowledge what all should I find out after mean and standard deviation. – Kerry Mar 04 '13 at 11:35
  • "I am not allowed to use any existing libraries. I should code on my own" rather suggests this is some kind of educational assignment; in which case do you not think you should *do it yourself*, using the material you've presuambly been learning from? – AakashM Mar 04 '13 at 12:18
  • http://stackoverflow.com/questions/9996327/using-a-naive-bayes-classifier-to-classify-tweets-some-problems/10163626#10163626 – etov Mar 11 '13 at 21:44

1 Answers1

0

Naive Bayes is useful to predict output out of your two networks(0,1) using independent parameters zip, city and others.

so you use "NaiveBayes Bernauli formula" to modulate and predict any case. i suggest you to read RCRAN Logistic Regresion formula-

https://cran.r-project.org/web/packages/HSAUR/.../Ch_logistic_regression_glm.pdf

Ram

Ram
  • 9
  • 2