Hey folks, I just released a new version of statsforecast (notes). The most important new feature is the MFLES model, developed by a community member which achieves a better score than many of our current models and can be substantially faster with just the target, however it also supports exogenous features and multiple seasonalities. I think it'd be great to make an announcement, please let me know if I should create a draft or similar.

Awesome! Thanks, yes, I think it would be great to make an announcement @José Morales. If you can write a draft, especially more about the MFLES method we can share that. I started a draft, but please feel free to change anything there.. https://docs.google.com/document/d/16Dat9seQfJk5C5WY9-0w01SfanyjwYMB23984QYmZro/edit

Thanks! I've added some more content, please let me know what you think

awesome @José Morales! congrats. these are very good results. i think we can start a lot of concersations about this.

Are we sharing this today?

I gave it another bit of an edit. StatsForecast release information ready to post on LinkedIn from my perspective, so just let me know when to post.

Who needs to say when we should post it?

@azul (she/her) (nixtla) @Max ok to post this today on LinkedIn?

I think we should go for it

thanks @José Morales @Tracy Teal! tomorrow will be publishing content related to the stable release and i'm worried about that the mfles release will lose some impact, perhaps we can publish it next monday? we have seen a lot of impact at the beginning of the week, wdyt?

Sounds good, next Monday it is!