Predicting the Election

[zak100]

Hi,

I am trying to create a Byesian network to predict about the election results. For this election there are some important issues (I would call them as nodes) like EV, CD19, BLM, and WSM. I can assign some values to them. And then we can have two candidates,  T and B, they belong to some party and these parties have workers. Now I have to arrange the nodes to indicate the dependencies and then assign conditional probabilities.

Kindly guide me how can I improve my approach and what values should I assign to conditional probabilities and how to estbalish the network.

 

Zulfi.

[PoetheProgrammer]

I’d break it down into individual states (you almost certainly need to regardless) and pull some polling data as a start to get probabilities.  You’ll also need data about each states views on the issues at hand in order to properly account for them otherwise you’re just guessing.  I would start with modeling an individual state (e.g. Texas) and building a backtester to verify results.  Once you get that down add a couple of other states to guarantee you aren’t overfitting and scale up from there.

[zak100]

Hi,

Thanks for your reply.

<You’ll also need data about each states views on the issues at hand in order to properly account for them otherwise you’re just guessing. >

I would search. Right now I have found general issues which I think are applicable to all states but of course there are issues like Economy, Climate Change, Migrants which can relate to other states.

< I would start with modeling an individual state (e.g. Texas) and building a backtester to verify results.>

This is what I wanted to do because I don't know how to create a network. Once I evaluate the results (wrong/right probabilities ) but with a correct nectwork, I think I can do that for other states.

Please guide me how to establish dependencies and how to assume values  for conditional probabilities table (CPT).

Zulfi. 

 

 

[zak100]

Hi,

I have created a bysian network using samiam tool. 

In the diagram, I have shown 3 major issues but let’s consider that there are 4 issues and  I have got following data related to Texas state:

C19=COVID19 = 38%

EJ= Economy and Job (28%)

R=Racism(9%)

HC = Health care = (9%)

Party Affiliations:

D= Democrat = 39%

Re =republican= 40%

N = neutral

There are total 17 million voters

Somebody please guide me how can I use the above data to predict the inclination of voters to one of the two candidates in a two party System where one of the candidates is the incumbent (i.e. I want to get an estimate of the actual number of voters inclined towards a candiddate).Please feel free to suggest the modifications. Please guide