search

Jindi0 / SAFE

59 stars 15 forks source link

How to get case_y_fn.npy？ #10

Open fovik1126 opened 2 years ago

fovik1126 commented 2 years ago

Hello author, I don't know how to get this y_fn, it should be 0 or1 in y_fn, does this also need to be embedded by the SIF tool?

Jindi0 commented 2 years ago

Hello, for the binary classification with labels "fake" and "real", we use 0 and 1 to represent them, respectively. They do not need to be embedded.

fovik1126 commented 2 years ago

Hello, could you please tell me whether you need image description embedded, if necessary, embedded length should be??

thinking024 commented 1 year ago

I'm a little confused, in your paper the label of fake news should be 1 and real news should be 0. Besides, the input y seems to be 2d ndarray in your code, which means the input label of each news should be a vector like [1, 0] or [0, 1] but not scalar

Jindi0 commented 1 year ago

Yiyuan,

For the binary classification, the output of the final fully connected layer must be a vector in size of 1x2. So we use oh-hot encoding to represent the two classes. It does not matter which on-hot code you use to represent label 1.

Jindi

From: Yiyuan Zhu @.> Sent: Monday, November 7, 2022 8:46 AM To: Jindi0/SAFE @.> Cc: Jindi0 @.>; Comment @.> Subject: Re: [Jindi0/SAFE] How to get case_y_fn.npy？ (Issue #10)

But in your paper the label of fake news should be 1 and real news should be 0. Besides, the input y seems to be 2d array in your code, which means the input label of news should be 2d array like [1, 0] or [0, 1]

— Reply to this email directly, view it on GitHubhttps://github.com/Jindi0/SAFE/issues/10#issuecomment-1305639935, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AP56TZO5W3NXRB6TSYAWCOLWHEB25ANCNFSM5ZWW6VYA. You are receiving this because you commented.Message ID: @.***>