Відео

It is totally useless wtf ??

❤❤❤ from Nigeria

I know this is really late but can anyone explain to me how I change the data on here for a more recent date and also what line of code is making it only show data for 2018?

plz dont invest in crypto especialy waqar zaka group or u will be destroying ur life

Great tutorial!

its not even close to the price.. Why didn't u normalize the data first?

Would a model need to be retrained for each individual share, or would it be able to make predictions about all shares in general after it had been trained on just one share?

Sorry dislike

It seems that you are poorly trained in correctly identifying buying and selling positions!!! Where in the market should you sell when the price rises?!!!!

Very insightful...

great stuff :)

great topic thanks

my code to train the model keeps loading but will not produce an output thus i cant finish the whole analysis. Does anyone know how i can fix this? The code: #Train the model model.fit(x_train, y_train, batch_size=64, epochs=1 )

Thank you informative

i dont understand cd C:\Python38-32 dose not work and the first one dose not work eaither

Thanks alot. great video.

So you make a 21 minutes video to show nothing ?

Thank you for your help! Great video :)

It is genuinely fascinating looking at this over three years later - Marriott has done pretty well, but all the others have been somewhere between meh and abysmal. Past performance is not a good indicator of future results!

Hi i'm writing from asia, trying to learn using your video. however i don't get the tesla csv file. can you include in your links.

found on tests with ticker 'SAP.DE' and other stocks at 5 min, 1h , 1d interval to get max accuracy in those intervals: 5 min: 99.55 % accuracy 1 h : 80.12 % 1 d : 94.46 % ------------------------------------- using notebook: type into cell: !pip install xgboost helped me to import xgboost as xgb (py) and import yfinance as yf data = yf.download(ticker, period='60d', interval='5m')

useless video, not even showing how this "python" is installed into cmd.

Hi, is there an update for this project coming soon by any chance? Seems like many people including myself are running into an issue with pandas_Datareader not functioning with yahoo api and yfinance is the new thing. Maybe you can help ?

There are so much remaining to learn, but this was some intresting stuff. I learned about LSTM(Long Short Term Memory), taking the data from web directly, preprocessing the data,training the data and also visualizing the data.

but how to use that on live?

why don't u imported pandas library ????

Thank you for the code walk-through on this graph. 💡Love it!

How can we scale this to show next 30 day prediction from today?

9:17 please let me know if I m wrong I think that X should not include target value, so it should have 9 columns instead of 10 and exclude y value. So I think that X should be ---> X = dataset[ : , 0:9] Is it correct ? Thank you!

That's just stupid shit, your model is just replicating past data with a lag, that's so stupid. How can this crap video have so many views?

I like your videos very good- size fonts is good because I use my mobile for watch vídeo everytime..

No train test split. This is the equivalent of giving the model the answer sheet to the test so you don’t get an accurate picture of model performance

Thank you!!

You saved my life, I will name my first born after you. Be blessed until the neighbors chicken lay gets in your house

May I please get a copy of the code?

This approach does not work. To see that you're deluding yourself into believing it does, let it predict the difference between two subsequent close prices instead of the next close price. These two problems are mathematically equivalent, yet you'll get nothing but noise. Why? Because you cannot predict prices. Traders don't predict prices. The next close is approximate to the previous one. So your model spits out the previous close + a random epsilon, it's not predicting anything.

after we had trained and test it how can we make a way so user can himself enter the values and predict the price? Thank you for your help

Sorry,I mean several c files

Hello,I have a Sourcecode with several cs files.How to do that?

I see your code to calculate Strategy return = strategy return * return ( normal return) I think it must be dependent

anyone share collab repo

Did u not do prediction using any ML algorithms

Nice video, but your strategy has look ahead bias as you use the entire dataset to construct your percentiles. This means that the percentile values take into account future price data that would not have been known at the time. One solution to try would be to calculate percentiles based on a rolling window of the past X days.

Thank you so much, very appreciate your tutorial everything clear and worked perfectly!

what risk measure are you implementing