Fake News Detection using LSTM in Tensorflow and Python

Поделиться
HTML-код
  • Опубликовано: 5 ноя 2024

Комментарии • 117

  • @GAURAVKUMAR-bc9vd
    @GAURAVKUMAR-bc9vd 3 года назад +5

    Hi sir, after a long time thanks for uploading video on You Tube, I was thinking that you won't be active on you tube now. You are my favorite teacher your content have helped me a lot.

  • @vidulakamat6564
    @vidulakamat6564 3 года назад +3

    Great content. Thank you. Do we need to handle imbalanced data before training the model ?

  • @test-qk1ql
    @test-qk1ql 3 года назад +2

    I always learned a lot from your videos! Do you think you can make a video to talk about how skip-gram can be used to predict the context for a given target word? Thanks a lot!

  • @kenmwangi7550
    @kenmwangi7550 10 месяцев назад

    Happy New Year @KGPTalkie, Very helpful and great content sir. Is there any other reason why the accuracy level would stop at 65% despite using the exact same code and dataset? I have tried to change the DIM and units but 65% is the best accuracy level I am getting with an average loss ratio of 65% as well. Kindly assist. Thanks

  • @youssefelagdah4401
    @youssefelagdah4401 10 месяцев назад

    Sir, I want to ask you a question: Why did you combine the title with the text? Isn’t it enough to rely on the text only and erase the title?
    How does the title effect the identification of real and fake news?

  • @ВасилийНоженькин
    @ВасилийНоженькин 7 месяцев назад +1

    print(w2v_model.wv)
    KeyedVectors
    But u have 231911 length
    Do you know what problem can that be?

  • @rabiafaheem7983
    @rabiafaheem7983 2 года назад

    which algorithm are you using in this tutorial?

  • @yasirabdulkareem9844
    @yasirabdulkareem9844 2 года назад

    Great work 👏 . I have question . If we don't use vec2wrd and we make trainable = true ( in embedding layer) for generating word embeddings , is this right ? Thank you

  • @20-415-SubikshaM
    @20-415-SubikshaM Год назад +4

    how does your model predict that covid news to be true when your dataset has no such data?

    • @orlando_kawaii
      @orlando_kawaii Год назад

      Ofc a trained ml model can predict it. Thats the whole goal, i.e to predict on unseen data

  • @piyushvyas2475
    @piyushvyas2475 2 года назад

    Very informative one. Do you have any video where you performed tuning the keras LSTM parameter for text classification task. Please share the vid link. Thank you.

  • @shadwar123
    @shadwar123 9 месяцев назад

    Please make a video how to deploy it as api to use in frontend .

  • @bharatchandra2512
    @bharatchandra2512 3 года назад +2

    Thanks for uploading video after long time sir, please provide the link for full code in blog. I've opened the link mentioned in the description but I'm not able to find the full code

  • @roseliemrollno-3353
    @roseliemrollno-3353 2 года назад

    Your explanation is very good.I also tried this. But I got same output for all input. I don't know how to fix this problem.plz help me sir

  • @MRAHNO-ul9yf
    @MRAHNO-ul9yf 3 года назад +1

    Great content bro it detects real time news ...

  • @zavier3436
    @zavier3436 3 года назад

    sir please make a video where you are makeing an website and deplyoning the project

  • @LightningBeast-em3dc
    @LightningBeast-em3dc 5 месяцев назад

    facing keyerrors, namely u.s is not found in dictionary

  • @mihirbhawsar3150
    @mihirbhawsar3150 3 года назад

    Please make a video on multivariate Lstm model

  • @abdelrahmanhegazy3095
    @abdelrahmanhegazy3095 3 года назад +1

    this dataset can predict any covid news ?

  • @rabiafaheem7983
    @rabiafaheem7983 2 года назад

    Please teach how to make frontend for this model

  • @neto212
    @neto212 2 года назад

    Thank you for the great content. Greetings from Brazil.

  • @notknown9307
    @notknown9307 3 года назад +2

    After a long time I saw you,how was your Holidays and what happened to your camera lens ?

    • @KGPTalkie
      @KGPTalkie  3 года назад

      Thank you so much for your love. I am good. I hope you are also. After lots of struggle finally they had to refund. Thanks a lot for your support.

    • @notknown9307
      @notknown9307 3 года назад +1

      @@KGPTalkie that's a good news Flipkart refunded😁😁
      Be safe be healthy because of second wave of covid19 in india:)

    • @KGPTalkie
      @KGPTalkie  3 года назад

      Thank you

  • @sadikaljarif9635
    @sadikaljarif9635 2 года назад

    is this word2vec follow cbow method or skip gram???

  • @mohansai8735
    @mohansai8735 2 года назад +1

    can you share that colab file so that we can refer

  • @raghvendrakaloo2719
    @raghvendrakaloo2719 2 года назад

    Can you tell how to attach this model in flask WEBAPP ?

  • @amitdutta3875
    @amitdutta3875 3 года назад +1

    Please upload more on natural language processing

  • @DevendraYadav-sf6ry
    @DevendraYadav-sf6ry 3 года назад +1

    understood almost everything but can anyone answer one doubt that is why did we used weight_matrix as np.zeroes was it necessary to create..

    • @KGPTalkie
      @KGPTalkie  3 года назад

      To avoid dynamic memory allocation. I created it with zeros before it's actual use.

    • @DevendraYadav-sf6ry
      @DevendraYadav-sf6ry 3 года назад

      @@KGPTalkie thanks alot...

  • @yashkumargoel7045
    @yashkumargoel7045 3 года назад

    Sir plz make a vd on multilabel news classification using nlp and deep learning

    • @KGPTalkie
      @KGPTalkie  3 года назад

      Already have one. Please see in nlp playlist

  • @bhumivyas7233
    @bhumivyas7233 3 года назад +3

    I am getting accuracy between 59-65% for the same code followed, what should be done to increase the accuracy?

    • @KGPTalkie
      @KGPTalkie  3 года назад

      you can change last units and layers

    • @bhumivyas7233
      @bhumivyas7233 3 года назад

      @@KGPTalkie What should i change in layers?

    • @LaxmiKantTiwarilaxmimerit
      @LaxmiKantTiwarilaxmimerit 3 года назад +1

      @@bhumivyas7233 You can increase or decrease LSTM layers. and you can also change the embedding dim. these are part of hyperparameter optimization.
      like model.add(LSTM(units=64)) or model.add(LSTM(units=256)) etc

    • @bhumivyas7233
      @bhumivyas7233 3 года назад

      @@LaxmiKantTiwarilaxmimerit I am not getting how to do what you are telling, please help.

    • @bhumivyas7233
      @bhumivyas7233 3 года назад

      Please help

  • @gaganashreekm3750
    @gaganashreekm3750 Год назад

    Sir !pip install spacy==2.2.3
    Is not installing in my laptop please help me with it

  • @arushigupta9510
    @arushigupta9510 Год назад

    isnt the model overfitting?

  • @krishnakumarvishal883
    @krishnakumarvishal883 2 года назад

    model.fit(X_train,y_train,validation_split=0.3,epochs=6) from that line it takes more time and it is disconnected and reconnecting again sir

    • @ismailazouz7494
      @ismailazouz7494 2 года назад

      i have the same problem did u find a solution ?

  • @sauhardyasatabdipradhan6476
    @sauhardyasatabdipradhan6476 2 года назад +1

    The cell in Google colab is not running. I have implemented according to the steps and also haven't got any error, but the cell is not running last line
    model.fit(X_train,y_train, validation_split=0.3,epochs=10)
    It shows message previous execution ended unsuccessfully
    Executed at unknown time
    I also restarted and run all the cells, checked every bit of code,but still this problem is happening.

    • @ismailazouz7494
      @ismailazouz7494 2 года назад

      I have the same problem did u find a solution for this problem

    • @gurvinder8723
      @gurvinder8723 2 года назад

      You need to be connected to the GPU.

  • @MRAHNO-ul9yf
    @MRAHNO-ul9yf 3 года назад

    Iam using Jupyter notebook..if I do in collab will the error stop

  • @AravindHan008
    @AravindHan008 3 года назад

    thanks for the lecture sir , if you provide the source code link that could be very helpful to tweak and learn

  • @SWAMIMISHRABEC
    @SWAMIMISHRABEC 3 года назад

    Did flipkart helped sir? very nice project

    • @KGPTalkie
      @KGPTalkie  3 года назад

      Flipkart didn't but you guys have helped me a lot.

  • @seo7854
    @seo7854 2 года назад

    Great video man :) thanks

  • @arushigupta9510
    @arushigupta9510 Год назад

    why did u take the character limit to be less than 120? Twitter character limit previously was 140 and now its 280

  • @gaganashreekm3750
    @gaganashreekm3750 Год назад

    Sir !pip install spacy==2.2.3

    • @gaganashreekm3750
      @gaganashreekm3750 Год назад

      Sir !pip install spacy== 2.2.3
      Is not installing is giving error please help me sir

  • @rabiafaheem7983
    @rabiafaheem7983 2 года назад

    i am getting error
    tokenizer = Tokenizer()
    tokenizer.fit_on_texts(X)
    method object is not iteratable

  • @MRAHNO-ul9yf
    @MRAHNO-ul9yf 3 года назад

    Can you tell me the version of tensorflow , numpy and python that you are using for running this model..

    • @KGPTalkie
      @KGPTalkie  3 года назад +1

      Use Google CoLab. There it was tensorflow 2.4

  • @heisenbergwhite5845
    @heisenbergwhite5845 3 года назад +1

    Really useful!
    Where can we get the source code?
    Thank you!

  • @SharathsooryaBCS
    @SharathsooryaBCS 3 года назад

    How can we hide the backend code since when user tries to find the fake vs real news then he should see only the gui and the code should be hidedm

  • @deepchoudhury2001
    @deepchoudhury2001 3 года назад +1

    Can't find the blog in the website please give the link

  • @upasanabhardwaj6140
    @upasanabhardwaj6140 3 года назад

    sir im unable to install wordcloud in jupyter notebook. anconda

  • @lokeshparab
    @lokeshparab 2 года назад

    I tried to code this But I can't able to train this in colab due to less memory how to proceed this

  • @upasanabhardwaj6140
    @upasanabhardwaj6140 3 года назад

    sir the dataset files are not downloadable? And from we get the source code of this model?

  • @deepikag3760
    @deepikag3760 3 года назад

    Hi sir can we create GUI for this?

  • @yassinehab6794
    @yassinehab6794 Год назад

    How can i get the code asap

  • @MRAHNO-ul9yf
    @MRAHNO-ul9yf 3 года назад

    Bro why I have this notimplementederror: saying cannot convert a symbolic Tensor(lstm_1/strided_slice:0) to a numpy array.

  • @chadhuvukoraaa9236
    @chadhuvukoraaa9236 2 года назад

    where should i get source code

  • @MRAHNO-ul9yf
    @MRAHNO-ul9yf 3 года назад

    Bro I have not implemented error in model.add(LSTM (unit=128))..please help

  • @muhammadnaeem4754
    @muhammadnaeem4754 3 года назад +1

    Please share the CSV files , i searched both the given website and your github there is no csv file for fake and real dataset,

    • @KGPTalkie
      @KGPTalkie  3 года назад

      github.com/laxmimerit/fake-real-news-dataset

    • @footballforfun2054
      @footballforfun2054 3 года назад

      @@KGPTalkie where can I get your code?

  • @krishnajangir1784
    @krishnajangir1784 3 года назад

    Can we run this on Jupyter Notebook?

    • @KGPTalkie
      @KGPTalkie  3 года назад

      Yes

    • @mohammedaldobai2415
      @mohammedaldobai2415 Год назад

      @@KGPTalkie l have implemented this code unfortunately I have got the 53% accuracy.could you plz help me to get a high accuracy like the video.

    • @KGPTalkie
      @KGPTalkie  Год назад

      Did you follow code as shown in video and using same data?

    • @mohammedaldobai2415
      @mohammedaldobai2415 Год назад

      @@KGPTalkie yes I follow the same code unfortunately I have got 53% accuracy. Using Jupyter Notebook. Please help me. I will be grateful to you so much

    • @KGPTalkie
      @KGPTalkie  Год назад

      please share your notebook to udemy@kgptalkie.com

  • @hkcs1206
    @hkcs1206 2 года назад

    model.fit(X_train,y_train,validation_split=0.3,epochs=6) i m getting errror here

    • @aryac845
      @aryac845 Год назад

      Me too. Did u find any solution?

  • @prateekrajput2646
    @prateekrajput2646 Год назад

    Thanks lord god bless you!

  • @LightningBeast-em3dc
    @LightningBeast-em3dc 6 месяцев назад

    are you intentionally not providing the code ?

  • @indukumari193
    @indukumari193 2 года назад

    hello sir could you share the code on my mail ...your blog page is not opening

  • @paniz1479
    @paniz1479 2 года назад

    1:13:41 at assigning the weights, I get an error: “word ‘u.s.’ not in vocabulary”
    Can you please help me?

    • @aryac845
      @aryac845 Год назад

      Iam getting same type of error. Did u find any solution?

  • @jithinkuriakose447
    @jithinkuriakose447 3 года назад

    Sir, am stuck @
    w2v_model = gensim.models.word2vec(sentences =×, size=DIM, windows =10)
    Error is.
    "Module object is not callable"
    Can you please help me out. Please
    Or please provide any of contact

  • @gnanasekar6468
    @gnanasekar6468 3 года назад

    Sir code link doesn't work

  • @dipeshsaili4468
    @dipeshsaili4468 3 года назад

    Where's the code sir?

  • @VibeWithMe3i
    @VibeWithMe3i 3 года назад

    Code ?

  • @pukingrainbows1609
    @pukingrainbows1609 2 года назад

    Does anyone have the source code of this?

  • @omsutariya02
    @omsutariya02 Год назад

    39:00 you didn't explaind the reason :(

  • @fiverr.mairaali26
    @fiverr.mairaali26 3 года назад

    At 1:14:54
    This error is coming "KeyError: "word 'ifax' not in vocabulary"
    Please help me solve this error.
    Thank you

    • @shailendrapasi3577
      @shailendrapasi3577 3 года назад +1

      for word, i in vocab.items():
      try: weight_matrix[i] = model.wv[word]
      except: print(word, 'not found')

    • @josyulapravallika4652
      @josyulapravallika4652 2 года назад

      hey can u plss tell me, did u find solution

    • @aryac845
      @aryac845 Год назад

      Iam getting same type of error. Did u find any solution?

  • @infotainmenta1828
    @infotainmenta1828 3 года назад

    1:15:31

  • @omsutariya02
    @omsutariya02 Год назад

    after 1:00:00, you didn't explained the code properly.

  • @TheAstralftw
    @TheAstralftw 2 года назад

    you didnt have proper data set to test how good is your model in real life. once sentence is not enough.

  • @MRAHNO-ul9yf
    @MRAHNO-ul9yf 3 года назад

    Can I use anaconda rather than Google collab