Project 4. Fake News Prediction using Machine Learning with Python | Machine Learning Projects

[@Siddhardhan]

Hi! You can join this Telegram group for regular updates about my videos: t.me/siddhardhan Thank you!

[@vishalkumarpal9897]

Where is link of hands on data science course ?? 12:02

[@vimal554]

Sir one doubt where is output of this project ☝🏻☝🏻☝🏻

[@antibioticsOfWorld]

@@vimal554 output itself is shown ..either a news is fake or true.. if u can elaborate ur doubt ..let me know.. i hope i can help u 😄

[@harissaeed5811]

Fake Supporters Detection how can we do this using ml ? need your help

[@durgabhavani5964]

@@vishalkumarpal9897 sir can u pls send fake job posting prediction document

[@harshaddeshmukh2935]

Hat's of to your teaching skills sir! As a beginner, this was my 1st ML project & I have completed it successfully & understood each & everything. Now, I am pretty sure that the pure golden content is hidden deeply in KZbin & as a viewer it's our responsibility to find it & share it with others who need it! Thanks a lot sir!

[@bantuman01]

It's so satisfying when someone blessed with talent such as yours is able to simplify a project enough for the average student to get a grasp of what the project is all bout. Thank you so much!

[@renusriPalaparthi]

Hii

[@mohdhammadkhan5570]

You have provided so much projects while the paid courses worth of 30000 provides 6-7 projects. Great work💯

[@Siddhardhan]

you are most welcome 😇

[@86tejasd44]

Your programming part is really skilled and logics are also good i was searching for these kinds of topics

[@mathavansg9227]

In case u have an error “EOF inside string etc.”, just wait until the train file finishes loading. It is pretty huge

[@dipakrajbhar2723]

Thank you so much

[@s.mth8352]

Thank you!! I was stuck here while running my code😅

[@sanjuraiker]

I need source code please

[@renusriPalaparthi]

Hey hi

[@akritisingh7240]

Hello... Can you tell me how to connect this code with frontend?

[@gameofcodes9832]

This series is really useful for beginners like me. If you could do a video on deploying these models for real-world use then it would be a great help.

[@Siddhardhan]

Sure thing! will do after deep learning course

[@gameofcodes9832]

@@Siddhardhan wow great. BTW did you already started deep learning series?

[@rn4341]

man, im new to this and I cant thank you enough for this material. I understood everything you did, but i still have trouble doing this on my own due to my lack of experience in the syntax and what modules to import :/

[@bushin.ashura]

Thanks a lot Siddhardhan. I will let my friends know about your channel.

[@devarpanatribedi1281]

Outstanding teaching style you possess sir, I can't thank you enough for this pure gold-like content. I appreciate your idea of bringing these project-related videos and making them understandable to all students. Thank you.

[@extremexplorer8930]

Awesome so clear Understanding of Machine Learning....Best Machine learning Channel on youtube where we can truly learn Others are there for Views only.

[@navneetgupta4669]

Nice tutorial. The best part is that the video doesn't have annoying ads.

[@Siddhardhan]

Thanks 😇

[@yohannesayana9456]

Your video is incredible. but is there a means to check the prediction by inserting textual news instead of numerical index?

[@huzaifaali5245]

most useful project as it helps me learns a lot . Thanks alot sir

[@bishalrajak8681]

Nicely explained . Can you please make a video regarding the mathematics behind Logistic regression in details

[@sezermezgil9304]

Great tutorial ! And i have a question while we were making predictive system we have used data from X_test and we said we are using this because the model have not seen these datas.But we also write a code ''X_test_prediction=model.predict(X_test)''.Didn't it see the test datas after we predicted that ? thx.

[@asimmubeen]

Your actual test output that you are predicting is in Y_test variable. Basically your X_test_predictions are predicted Y-values. I hope it will help.

[@dikshagupta3276]

Hi siddhardhan nice explanation Pls make some end to end project where you discuss on eda and features engineering, features selection ,remove outlier I want this type of project pls make these type of project

[@sejalanand23]

Really appreciate your explaining each and every line. Thank you for your time.

[@Siddhardhan]

My pleasure!😇

[@PenAndSpecs007]

Another amazing project sir! Loved it.

[@sonamraj5323]

Thank you so much for all your efforts sir ,it means alot...

[@followmanish]

thanks for the tutorial. but can you suggest how can we related the model with realtime world. like if I pass any news title to it, it should return the Real/False.

[@anasali8865]

Well done! Thank you! I have a question. Is it possible to figure out if one single news ist real or fake? In other words, I want to know if a one item of news is real or fake.

[@sabarikrishna4900]

Thank you so much sir, For providing us such good videos in the field of Ai & ML. All your videos helped me a lot, my interest in this field has boosted a lot since watching your videos. Keep posting more videos like this sir you are inspiring us a lot 🥰

[@mahakgarg3152]

My file is showing its too large to display .What can I do now?

[@siripireddyvaibhav4615]

for unicodedecode error for reading csv use this: news_dataset = pd.read_csv('train.csv', encoding="unicode_escape",error_bad_lines=False)

[@devanshprataptiwari5804]

thanks a lot buddy....wasted more than an hour but got resolved by you

[@sweety6734]

It is showing parserError how to solve

[@aslah4188]

​@@sweety6734 same how to solve?

[@tohando]

Nicely done! I like the recap at the end of the video. Keep up the good work!

[@Siddhardhan]

Thank you so much 😇

[@antibioticsOfWorld]

You are just superb explaining each and everything properly 🙌🙏 thank you

[@Siddhardhan]

Thanks and welcome😇

[@antibioticsOfWorld]

@@Siddhardhan I am doing my masters in Queen mary London and i was just surfing random videos ..even though i knew all the concepts in this i didn't feel like skipping a bit.. truly well explained..i have also sent u the LinkedIn request ! U will have bright future 😄

[@tanyanagpal6554]

Really liked the video and that recap part is very nice... I wanted to ask one thing: Can we show a graph of fake and real news? If yes, then please tell me the procedure for it! Thank you !! in advance

[@Siddhardhan]

thanks a lot 😇 go through this to plot the data in logistic regression: stackoverflow.com/questions/46085762/sklearn-logistic-regression-plotting-probability-curve-graph

[@ganeshsubramanian6217]

Hi Siddharth, this is really a good content. Does the video quality reduces when you upload thru KZbin?

[@premswaroop4337]

sir i really love the way u explain each and everything in detail, thank u

[@behappyvp4956]

Thank you very much. Wish you all the best.

[@gammetube3976]

Thank u a lot! Sorry show us how we apply part of speech tagging(POS) to hate speech detection using machine learning

[@RajeshmayBurman]

Bro i have used this dataset and applied random forest everything is perfect accuracy precision but ig its overfited any outside news its displaying as fake ??????

[@ymachannel9150]

if u are facing error in news_dataset['content'] = news_dataset['content'].apply(stemming) it's likely u did not add .stem at the back of port_stem so use this stemmed_content = [port_stem.stem(word) for word in stemmed_content if word not in stopwords.words("english")]

[@ankurtripathi8305]

Thank you for the video !! I have a question that: 1) I f I have a news (New Query) and I want to predict that is fake or real , How can I do with this model? Regards Ankur

[@JaswanthkumarPalavalasa]

Take it as input , transform the data to Tf-idf verctorizer using transform function and then proceed with same steps.

[@santoshr1708]

Thank you so much brother. Loved this one. You taught better than my trainer.

[@ShrishtiSwaraj-p8i]

Thankyou so much Sir. Helped me a lot.

[@naveenarameshvardhinij3526]

Hi, Can we able to put the exact news text and find the prediction rather than using index in predective system.if so can you explain?

[@paulubakanma0]

Thank you for this amazing video. I hope to use it for my project defense

[@miranbaban9554]

Dear, I have a question here. So how i can predict for a piece of new news, i mean it should be predicted directly without setting the label from ourselves. so how we can predict that.

[@vaibhavkirtankar5336]

Thanks a lot for this amazing project . i was checking code and first time you split data "X = news_dataset.drop(columns='label', axis=1)" then you did'nt use it .Was it necessary ?

[@saritha3957]

Best video for Beginners

[@charles007able]

Great work @Siddhardhan, can you suggest steps to write a simple rule based news recommendation system (Popularity & Content based) with summary metrics (like page views, timestamp etc) from analytics as inputs. Much appreciated!!

[@Siddhardhan]

Hi! you can follow the same procedure. You just need a dataset that contains details about popularity score.

[@charles007able]

@@Siddhardhan sure will do that, have learned a lot. Thanks a ton!

[@Siddhardhan]

Thanks a lot for your positive words 😇

[@ΓιώργοςΒαμβακούσης-ρ1τ]

Good job Siddhardhan!

[@Siddhardhan]

thanks a lot 😇

[@sanjayrao8406]

Hey there! Really grateful for you for coming up with this playlist of projects. I want to know how to extend the same model we used here to predict the news is fake or not, where we input the data outside of the dataset.

[@Siddhardhan]

Hi! Thanks a lot

[@Siddhardhan]

You can build a predictive system. Refer my "Project 1. Rock vs Mine Prediction". I have shown how you can build a predictive system. You just need to do some changes.

[@sanjayrao8406]

Okay say if I was able to come up with that part... How shall I find the data to input ?

[@Siddhardhan]

You need to give the details about the news title & the author. This will be your input data. Because those are the features we used. You can also train your model with description of the news.

[@sanjayrao8406]

Okayy will let you know if that works. Thanks for the help.

[@rishi-e3h]

in case anyone gets error in stemming function or the loc news_dataset['content'] = news_dataset['content'].apply(stemming) use the below code just before stemming function (it worked for me) news_dataset['content']=news_dataset['content'].apply(str)

[@samuelpaul5923]

Actual accuracy parameter is like this "accuracy_score(y_true, y_pred)". But in your video, you are giving the predicted value in y_true secion and true value in y_pred section. Please check on it and let me know if am wrong

[@Siddhardhan]

Yeah. You are right. I made a mistake there. But the value won't change much...

[@nandinigulati5704]

such good videos. Quality content. Thank you

[@adityamallick8867]

I am having trouble applying the Logistic Regression model on the test data that is provided on the Keggle page along with the training data given. How should I go about doing that??

[@anubhavjoshi3105]

thanks soo muxh for this awsm explaination

[@youtubelegends4499]

Please tell me feasibility study (economic, technical, operational) for This project (fake news prediction)... Please 🥺🙏 reply as soon as possible

[@gourangpathak4443]

Amazing Project

[@MohammadKumail-pl8pf]

can you please try to make some projects with tensorflow and keras . as I also want to learn those libraries

[@nothingmatters413]

Thanx man this really helped me 😭❤️

[@sharath957]

Good contents in your channel. Just subscribed to your channel and looking forward for more hands one projects.

[@Siddhardhan]

thanks a lot

[@aminullahyousufi8142]

Thank you it was really helpful

[@MAUMITABISWASRA]

Firstly, I want to say that your video is really very informative. Still I have few questions. 1.What does the testing data actually contain? I mean if it contains 20% of the training data, then why it is called the testing data and why it is used? Please if you can explain the whole concept of testing data. 2.# accuracy score on the training data X_train_prediction = model.predict(X_train) training_data_accuracy = accuracy_score(X_train_prediction, Y_train) What is basically this predict? Explain these two lines please. 3. Lastly i wanted to ask, that can't we feed any latest real time news into this model and get the output as real or fake? I've mailed you the same so if you can reply, it would be great, thankyou...

[@Siddhardhan]

1. it's not 20% of training data. it's 20% of the entire data picked randomly. 2. it predicts whether the news is real or fake. so it basically prints the labels 0 or 1. try printing X_train_prediction. you can understand it. we are comparing the prediction made my our model and the original labels using accuracy_score 3. in my other videos I have explained how to build a predictive system. it's basically the same procedure. you can refer that. with that predictive system, you can find whether a new news is real or fake.

[@MAUMITABISWASRA]

@@Siddhardhan Thankyou so much sir....

[@karanshethia3560]

@@Siddhardhan can you please help me with the 3rd point? . We'll input the news in the form of a text. Then do we need to first perform stemming and then convert it into numerical form using Tf-idf vectorizer and then feed it into the system?

[@ankit.thakre27]

That stemming function... No error occurred But I select wrong ^ symbol What that cause... Every English later erased and symbols left.... So whoever going to try this project... Please print the content after applying stemming. To verify And thanks siddhardhan for sharing colab file❤

[@prakhargarg4166]

what are the words, specific to fake comments. Like in sentiment analysis, we can say "good" and "bad" are some of the basic words. But, in this task, how will it distinguish

[@nishantchaudhary7528]

If we are getting new data/news, we have to convert that data through all the process before inputting in our model??

[@jarvis9263]

In case u have an error “EOF inside string etc.”, just wait until the train file finishes loading. It is pretty huge this is a coped comment, i hope it will be helpful and please copy this and send, so others can find it, many of us are facing the same problem

[@dora_2002]

hello sir. I am getting an error in the line -news_dataset['content'] = news_dataset['content'].apply(stemming).what to do sir??

[@sethukarasisethu6491]

Thanks a lot sir🌸

[@namansethi1767]

Thanks bro for all these projects

[@Siddhardhan]

you're most welcome 😇

[@prasantachanda7443]

Bro thank you so much for all these projects and Is there any other medium to contact you i need some help in my project??

[@jaiagnihotri5076]

great video man

[@akhilaarvapalli]

Hi sir while running model.fit(x_train,y_train) output it showing LogisticRegression() In this way empty can you tell my error

[@Siddhardhan]

no. it's fine. it's not an error

[@akhilaarvapalli]

@@Siddhardhan tq sir

[@charanpoojary4804]

THANK YOU VERY MUCH SIR

[@yasirabdulkareem9844]

Thanks 😊, just question, if we have unlabeled dataset , what we do in this case ? How can we label it ?

[@rohidaskhandge2548]

pd.read_csv( ' file path ', header= None )

[@gauravfamily2209]

great. But include data visualization of dataset in the project.

[@sezermezgil9304]

Instead of filling with empty string can we just drop those columns which is null ?

[@abc-yu3wd]

thank u so much sir! Sir can u pls tell how to deploy this?

[@zainqureshi7593]

hi hope you are fine how much time it will take the excution of only text data and is there any change required to excute only the text data?

[@abhishekwaghmare2480]

hello sir, when I try to download the dataset from link that you provided, it asks me to accept competition's rules before downloding file. What should I do?

[@Tony-vo9ok]

Amazing!

[@kanikagoyal6095]

hii @siddhardhan ....how to implement SVM and ANN in this project?

[@rajdeepsinhsisodiya2699]

thank you so much your video helped me a lot

[@Siddhardhan]

Glad it helped!😇

[@jesusalbertoargelramos8658]

Thanks bro!

[@shamimrahmansam]

Could I use it as a data science project in my cv?

[@shivaninaeck2524]

hello sir. Nice explanation. How can i use this to detect fake news for a mobile app on visual studio?

[@azizulpatni7373]

Which Machine Learning Algorithm is used in this project???

[@piyushsharma7956]

sir while data preprocessing an error keeps coming in the file downloaded from kaggle, ParserError: Error tokenizing data. C error: EOF inside string starting at row 1259 what to do? i have done various online solutions from stackoverflow but they arent helping

[@srcreates7757]

Is this problem solved or not

[@vancedaccount562]

If I enter latest news headlines from the internet will it be able to predict real or fake news. My college professor is asking for this feature. Please help me

[@hawaabasheer5239]

Where is the title column?why we didn’t feed it into model ?

[@whatsstatusvideo2201]

PC to an external monitor and I'm watcNice tutorialng tNice tutorials video on one screen and working through exactly what Michael is doing and it is a big

[@sandipansarkar9211]

finished watching

[@sandipansarkar9211]

hello I am unable to import the data set due to "utf-8".encoding issuePlease provide the exact code for that.Nowhere it is mentioned

[@zhrafin2578]

same

[@samiyusufzai9248]

Please post video on ML project for Telecom Churn Prediction,Thanks

[@dayobalogun5248]

Thank you for this video but what happens when our dataset is not labelled

[@akritisingh7240]

Can you tell how can i attach this code to frontend???

[@ayushmaansharma4075]

Did you get the solution?

[@manikyamsai3667]

What are the modules are there in this project.?

[@ddesmondeinni7758]

Hello Siddhardhan, how can I deploy this particular Fake news prediction code. Trying really hard nothing seems to work.

[@ayushiuniyal6242]

Can someone pleasee tell what are prerequisite before starting this project..like do i need to know all about machine learning..or some specific topics?..🙂

[@lillol3245]

Is it currently possible to make the prediction accuracy very close to 100% in context of fake news?

[@mariaflorenciadelvallegala5147]

Me gustaria saber lo mismo, porque tengo entendido que podria llegar a ser sobreajuste. Muy bueno el video!

[@subzero4787]

I have confusion whether we should use logistic regression model or simple vector model... Can we use simple vector model for this problem?

[@abhishekwaghmare2480]

Yes you can use it but if your accuracy score comes less then check for another model... I think you should go for RandomForest model. It gives better accuracy

[@subzero4787]

@@abhishekwaghmare2480 so we can randomly use any method isn't it

[@abhishekwaghmare2480]

@@subzero4787 yes you can but you have to choose according to your data otherwise your accuracy will fall .

[@nomanalikhan9998]

brother i want to ask something that we used author and title of the news but we did not use the text of the news it could be important for us that someone is using the real the author name and title but the text inside that news is fake.

[@santoshpaul934]

Sir, when i fit the logistic regression model the error "Unknown label type: 'unknown'" is coming. I stuck at that point. What should i do.

[@Shaik_Fathima]

Tqsm sir

[@sauravsuman9919]

bhaiya i am using ur videos for ml , please help me .it will help me a lot

[@86tejasd44]

sir but the thing is how to develop an app for this with frontend application lilke to give the random news for the project review

[@ShaileshKumar-db2mk]

How we can modify this project to giving title(text) of the News as input and the model give it is fake or not fake news?

[@the_deSo]

Take another variable and put the news there as string then vectorize 49:25 that to another variable, then again feed it in x_new