Hello, if you have any need, please feel free to consult us, this is my wechat: wx91due
Where {X1– Xm} are the features that are used for identifying the fake news, while {W1 to Wm } is the weighting of the features for all Xϵ {0|1} where 0 = absent, and 1= present, and Wϵ ]0.. 1].
Your group is required to design this Fake New identification tool including to study and determine the set of features {X1 – Xm }, and map with the W weighting. With Ф(.) = X1W1 + X2W2 + …. XmWm , you need to design the activation function to be the boundary and identify whether the tested news is the fake new or not. For this function, it suggestively apply the sigmoid function that will scale y between the [0-1]. The formula of the sigmoid function is: Sigmd(x) = 1/(1 + e-x )
A snapshot of your model
Suppose you define 5 features for identify fake news, which are [F1, F2, F3, F4, F5], with the weighting is [0.3, 0.5, 0.6, 0.1, 0.8]. In a particular case, the present of the features are [0,0,1,1,1], therefore the Ф(.) in your designed model is 0.6 + 0.1 + 0.8 =1.5, then Sigmd(1.5) = 0.81, it seems true (for 0.5 boundary)