New account since lemmyrs.org went down, other @Deebsters are available.

  • 0 Posts
  • 13 Comments
Joined 2 years ago
cake
Cake day: October 16th, 2023

help-circle
  • The duplicate content thing is kinda impossible to solve perfectly. Some people will tell you it’s a feature, and it can be interesting to see the different instances’ comment sections (especially after moderation), but yeah it can be annoying to have your feed dominated by a few stories.

    The default web front-end will merge crossposts, but won’t if they’re multiple posts to the same URL. I think some of the apps do have that deduplication as a feature, but I couldn’t tell you which.

    I remember the same problem from my Reddit days, but there wasn’t generally so many similar, overlapping communities.


  • From the Lemmy docs:

    • Active (default): Calculates a rank based on the score and time of the latest comment, with decay over time
    • Hot: Like active, but uses time when the post was published

    My default is set to

    • Scaled: Like hot, but gives a boost to less active communities

    This is the newest sorting option, I think, and it helps me not miss posts from the smaller comms - particularly ones where people are asking a question and there’s been no engagement. Ideally I’d like to have Mastodon-style lists so I could have “quiet comms” or something and check them all every so often.

    I will switch to new or top 6h/24h if I’ve been on recently and just want to see what’s fresh. Top all time or 1y if I’m looking at new-to-me comms so I can see what type of thing to expect from it.


  • Deebster@programming.devtoAsklemmy@lemmy.mlI'm new to lemmy. Any tips or advice?
    link
    fedilink
    English
    arrow-up
    19
    arrow-down
    2
    ·
    edit-2
    6 days ago

    There’s no algorithm here*, so use the different sorting options (for both posts and comments), as well as setting your favourite as default once you see what works for you.

    * the different sort options are of course algorithms, but I mean there’s no automatic, manipulative system like YouTube’s “The Algorithm”, Facebook, TikTok, etc.

    Voting doesn’t tune your algorithm, so I’d say only use downvoting for things that are low quality, trolling, in the wrong sub, duplicate posts, etc. Your votes aren’t private, by the way - although Lemmy itself doesn’t display voters’ names, that info is in every server’s database, and some other software in the Fediverse does show them.

    There are quite a few apps available, I like Voyager on Android and I stick to the default website on my computer.






  • Deebster@programming.devtoTechnology@lemmy.world*Permanently Deleted*
    link
    fedilink
    English
    arrow-up
    28
    arrow-down
    1
    ·
    edit-2
    27 days ago

    I found his paper: https://iopscience.iop.org/article/10.3847/1538-3881/ad7fe6 (no paywall 😃)

    From the intro:

    VARnet leverages a one-dimensional wavelet decomposition in order to minimize the impact of spurious data on the analysis, and a novel modification to the discrete Fourier transform (DFT) to quickly detect periodicity and extract features of the time series. VARnet integrates these analyses into a type prediction for the source by leveraging machine learning, primarily CNN.

    They start with some good old fashioned signal processing, before feeding the result into a neutral net. The NN was trained on synthetic data.

    FC = Fully Connected layer, so they’re mixing FC with mostly convolutional layers in their NN. I haven’t read the whole paper, I’m happy to be corrected.