AI Basics: Accuracy, Epochs, Learning Rate, Batch Size and Loss

Поділитися
Вставка
  • Опубліковано 25 сер 2024

КОМЕНТАРІ • 12

  • @ohoudid
    @ohoudid 2 роки тому +2

    يعطيك العافية

  • @1hello283
    @1hello283 2 роки тому +2

    haha.. I'm watching your video before I sleep. It's good to me to know the meaning of elements of the model because I am not native English language and you offer clear images and text to help me understand them.😊👋

  • @UsubbirAhmed
    @UsubbirAhmed 2 роки тому +2

    Excellent video.
    Crisp and clear.

  • @shadyayman851
    @shadyayman851 5 місяців тому +1

    great explanation
    clear & to the goal

  • @phuonglethithanh8498
    @phuonglethithanh8498 Рік тому +1

    Excellently explained video 🙌

  • @py6850
    @py6850 2 роки тому +1

    Very understandable, thanks you friend !

  • @fouadhellal5346
    @fouadhellal5346 5 місяців тому

    Thank you for your explanatiom, may i know the interface you're using to see the graphs ..etc ?

  • @frankocean5228
    @frankocean5228 Рік тому

    hi, what's the link for the previous video ?? couldn't find it. (Thank you for the detailed explanations 🙌🙌)

  • @GohOnLeeds
    @GohOnLeeds 9 місяців тому

    you don't put the day or the task in the video titles. Makes them almost impossible for us/youtube to find the last one/ next one ... not good :-)

  • @erwanerwan6196
    @erwanerwan6196 Рік тому

    hello sir !!!! i want 80% accuracy and more ...but in my images classification i can only reach 71%...... what is the issue ? thanks

    • @danielhermawan4054
      @danielhermawan4054 Рік тому +1

      Not enough training data? perhaps you should do some data augmentation to get more training data, other wise your data might be skewed (I rarely see this), there are a lot of approach to deal with low accuracy model, maybe also try using different model, add more layer, optimizer, error parameter, et cetera

  • @Abolhesh3354
    @Abolhesh3354 2 роки тому +1

    Thank you for your simple and clear presentation! Can I get your email address please. Much appreciated!