# Algorithm for weather prediction

I am trying to build a weather prediction app using c#. I am not a stats major and i am trying to understand which simple algorithm can be used to predict temperature and rain fall.

I have gathered temperature and rainfall data of fifteen years. I am currently looking into interpolation and curve fitting. I was hoping someone could provide me some guidance so i could get started.

For me, the implementation of an algorithm is more important than the accuracy of the prediction. I will be graded on the basis of implementation and manipulation of data.

• I'm extremely curious: what course is this that they would like you to predict weather but provide you no statistical background? Apr 16, 2016 at 7:11
• It is a course in computing and I selected advanced databases module. I selected this topic for my final year project. A more appropriate project would be student management system or something of that sort but I chose this particular topic. Of course, at the time, I didn't know what the hell I was doing and I already submitted the project plan for this. Apr 16, 2016 at 7:18

I can't help but give a very generic answer. A lot depends on what kind of sophistication is required; Is it a school project or some critical application. And further, a lot of time (weeks at least) can go into understanding the basic nuances of time series/forecasting. If you do not have the background in stats/analytics, you might overshoot your deadlines. I would suggest plugging-in one of these. If you really want to add forecasting to your forte, Rob J Hyndman's generous blogs, book and codes(in R) can be good starting point.

• It is a school project and the main objective is to apply a data-mining algorithm to predict temperature and rainfall. The accuracy of the predicted data, i imagine, is not as important as the implementation of the algorithm. The programming and data-manipulation is what i will be graded on. This is the reason i was looking into curve-fitting as it seemed to be do-able in my deadline. But I am not sure as i am new to stats. I have only taken up stats in my A-levels. I will update my question as well. Apr 16, 2016 at 7:03
• cool. I checked out the web to see if c# has time-series/forecasting libraries. It has :) (found this extremeoptimization.com/QuickStart/CSharp/ArimaModels.aspx ) this is not what you should directly go after. It just shows that if this exists other models (in free packages) should exists as well. read up the basic from the Hyndmans book though. Apr 16, 2016 at 7:14