Different implementation from the airxv paper

[Yeongtae]

Implementation different from airxv paper
https://arxiv.org/pdf/2410.12206
Abnormality Forecasting: Time Series Anomaly Prediction via
Future Context Modeling

I realize that this repository has a different title than the airxiv link.
But why is that link here?

[Yeongtae]

Arxiv paper indicates this repo is official.

But, I could not find these loss in your implementation.

[Yeongtae]

@zhaosinong

[Yeongtae]

FCM/exp/exp_long_term_forecasting.py

Lines 228 to 243 in 67e4c84

	series_loss += (torch.mean(self.my_kl_loss(series[u], (
	prior[u] / torch.unsqueeze(torch.sum(prior[u], dim=-1), dim=-1).repeat(1, 1, 1,
	self.args.seq_len)).detach())) + torch.mean(
	self.my_kl_loss((prior[u] / torch.unsqueeze(torch.sum(prior[u], dim=-1), dim=-1).repeat(1, 1, 1,
	self.args.seq_len)).detach(),
	series[u])))
	prior_loss += (torch.mean(self.my_kl_loss(
	(prior[u] / torch.unsqueeze(torch.sum(prior[u], dim=-1), dim=-1).repeat(1, 1, 1,
	self.args.seq_len)),
	series[u].detach())) + torch.mean(
	self.my_kl_loss(series[u].detach(), (
	prior[u] / torch.unsqueeze(torch.sum(prior[u], dim=-1), dim=-1).repeat(1, 1, 1,
	self.args.seq_len)))))
	series_loss = series_loss / len(prior)
	prior_loss = prior_loss / len(prior)

is Series L_C?
is prior L_det?

These are not MSE loss.