Nina maswali kadhaa,
1. Umetuambia tu kuwa hiyo algorithm itaapply kwenye physical land resources and environmental issues. This is too general. Naomba uniambie the problem definition. Yani what problem are you trying to solve? na in what scale (very large/medium/small)
2. Nimeuliza kwa sababu implementations za NN, SVMs and fuzzy logic will greatly influence the programming language you want to pick, naona wewe bado hujadecide on which one, but teari umedecide on the language. For example, training/testing data zinapoongezeka, SVMs become slow and consume very high memory kwa sababu zinatumia quadratic programming, hata NNs zinapata same problems with large increase of data but not as slow as SVMs. Sasa hapa pia inategemea na type of data na variables zako kama ni nyingi au la na to what scale you are planning to apply them.
Pia umesema unataka kudevelope Statistical software based on Bayesian Probability( Bayes Theorem), then use naive bayes, or regression, why jump to SVMs and NNs which are non-probabilistic? pia lazima uangalie kama unadeal na linear or non linear data. Labda kama ulisema bayesian probability to indicate using uncertainity to tackle problems, hapa ntakuelewa.
Kwa hiyo, cha msingi ni kucheki utakuwa unadeal na data aina gani? Then utatumia algorithm (method) gani? sasa baada ya hapo uje kwenye language gani ya kuimplement model yako.
Swali la mwisho, unatumia existing algorithms to meet your need, au unadevelope from scratch?