153 - Artificial Neural Networks - Explanation for those who understand linear regression

Поделиться
HTML-код
  • Опубликовано: 27 янв 2025

Комментарии • 46

  • @hemanthchenga5671
    @hemanthchenga5671 2 года назад +2

    one of the best RUclips channel to learn deep learning, kudos to you sir.

  • @rshaikh05
    @rshaikh05 Год назад

    excellent sreeni. I wish i had come to your channel before. now there is so much content that its overwhelming! but your explanation is so good with short videos covering only few concepts at a time and that makes it all manageable...

  • @srikanthpvr4161
    @srikanthpvr4161 3 года назад

    Excellent info which evry new learner should know how classicl machine learning to be used in NeuralNets.Presentation is superb and crisp

  • @brunospfc8511
    @brunospfc8511 2 года назад +1

    Great video, i'll recommend , and watch again

  • @manasranjanpanda9859
    @manasranjanpanda9859 3 года назад

    Thank you so much Sir, Indeed you’re a great teacher and being a PhD scholar in Hydrology I have attended machine learning course to use it in my research work but I am so grateful to you because you explain everything in such a way that a layman can easily understand...

    • @DigitalSreeni
      @DigitalSreeni  3 года назад +1

      I was layman once, and still am... so I feel your pain. This helps me structure my content accordingly.

  • @flanker6212
    @flanker6212 3 года назад

    Thanks!

    • @DigitalSreeni
      @DigitalSreeni  3 года назад

      Thank you very much for your support Flanker6. I really appreciate it!

  • @vzinko
    @vzinko 11 месяцев назад +1

    just to be clear and for others' benefit, standard relu is not used in the output layer for nonnegative regression problems

  • @bheemannanayak6728
    @bheemannanayak6728 2 года назад

    Hello sir, i have only one variable is there that is Age Dependency ratio data so how to do prediction for next 10 years using Artificial Neural Network.

  • @sumitmaitra
    @sumitmaitra 3 года назад

    Very nice video.

  • @dilipgyawali1776
    @dilipgyawali1776 3 года назад +1

    for this tutorial... which one is the dataset?

  • @samarafroz9852
    @samarafroz9852 4 года назад

    Best tutorial of ANN

  • @larrysummer2015
    @larrysummer2015 3 года назад +1

    Great video.. One question what activation function are you using in the last or output layer?

    • @larrysummer2015
      @larrysummer2015 3 года назад

      Please suppose I got negatives in my target.. What should be my activation function

  • @abderrahmaneherbadji5478
    @abderrahmaneherbadji5478 4 года назад

    Hello Sreeni.
    Please let me know, how we can use SVM classifier to classify deep features (CNN) instead of softmax function.
    Thank you.

    • @DigitalSreeni
      @DigitalSreeni  4 года назад +1

      I’ll record a video on the topic, stay tuned.

  • @raihananuralya8330
    @raihananuralya8330 2 года назад

    Hi sir. I still confuse about using ANN on linear regression data (on x and y). Better i use the code on this video or at your '155' video?
    Thank you!

  • @samarafroz9644
    @samarafroz9644 4 года назад

    Your videos r literally best sir

  • @charchikasinha
    @charchikasinha 3 года назад

    Very thorough explanation! Did have that aha moment :)

  • @aomo5293
    @aomo5293 2 года назад

    HI,
    For the mse you have used: np.mean(y_test-y_predicted)**2, I think this is error;
    mse = np.mean((y_test-y_predicted)**2)
    Thank you

  • @RajeshSharma-bd5zo
    @RajeshSharma-bd5zo 4 года назад

    Amazingly explained!!

  • @frankdearr2772
    @frankdearr2772 2 года назад

    Hi, usefull, perfect informations about how it works. great lesson, thanks a lot :)

  • @gh.g1084
    @gh.g1084 2 года назад

    Thank you!

  • @abderrahmaneherbadji5478
    @abderrahmaneherbadji5478 4 года назад

    I appreciate your efforts

  • @bratsummer1980
    @bratsummer1980 3 года назад

    Sir you are genius

  • @prasadbhadane608
    @prasadbhadane608 3 года назад

    Thank you so much sir!!

  • @mohamedramzyibrahim6375
    @mohamedramzyibrahim6375 4 года назад +1

    Hello, thank you for your efforts. I want to ask you to explain RCNN, faster RCNN, Faster-RCNN, RNN and LSTM with their implementations. Thanks very much for your help and i hope you could answer my comment and explain them. Your explanation and code snippets helped me in alot of understanding and going deeper in cnn and feature extraction methods.

    • @DigitalSreeni
      @DigitalSreeni  4 года назад

      I'll do LSTM soon. I have to find time to experiment with RCNN on real datasets before I can record videos. I do not like to use standard datasets, you can do a google search to find many such examples. For custom datasets I do not have labeled data in COCO format. This is the reason why I haven't done RCNN tutorials.

  • @efremyohannes2334
    @efremyohannes2334 4 года назад

    Hello, your video helped me a lot, thank you. Could you please work video on Bi-directional RNN-LSTM for the text line recognition?

    • @DigitalSreeni
      @DigitalSreeni  4 года назад

      A couple of LSTM videos are coming, so please stay tuned.

  • @nguyenhongthai2970
    @nguyenhongthai2970 4 года назад

    Thank you for sharing, nice tutorial
    Sir, can you make a video about medical hyperspectral images?

    • @DigitalSreeni
      @DigitalSreeni  4 года назад +1

      Never worked with medical hyperspectral images. Where can I learn more about them?

    • @nguyenhongthai2970
      @nguyenhongthai2970 4 года назад

      @@DigitalSreeni sir, it's just my suggestion. I learned a lot through your channel. Probably from science papers, I think the data type of hyperspectral images is the same as RGB images, instead of 3 channels, now we have hundreds of channels as wavelength bands.

  • @motriz-industrial6846
    @motriz-industrial6846 2 года назад

    Very clearly explained! Thank you!
    Been trying to find the equation for this data:
    drive.google.com/drive/folders/1LpLXfQLYWur0I29MmJBbm01f5fWsBcxL?usp=sharing
    I wonder if there is something that today’s technology (software) offers to turn this data into a model.
    Can neural networks do it?

    • @DigitalSreeni
      @DigitalSreeni  2 года назад +1

      Not sure what you are trying to achieve here. The data shows some numbers jumping up and down. There is no visible pattern in the numbers, seem to be rather random. Not sure if there is any periodicity but you can try any of the time series forecasting approaches, like ARIMA.

    • @motriz-industrial6846
      @motriz-industrial6846 2 года назад

      Thank you very much for taking the time to respond.
      Those are shaft rotations (in Degrees) that move a mechanical arm through a stepper motor. An expensive PLC runs the required motion using that data without problems, but if I could find the equation, I could use a very inexpensive PLC. I will take a closer look at ARIMA. Thanks again for everything.