162 - An introduction to time series forecasting - Part 2 Exploring data using python

[@juanpablomuniesaprados374]

Congratulations on the magnificent work!! Perfect balance between depth and relevance of each topic. Proper language and impeccable timing! It was a pleasure to follow your class!!

[@awaisahmad5908]

The best teacher. Everything i have learned from you. My University had never provided my any content or any help. Thank You !

[@DigitalSreeni]

Happy to hear that!

[@krishraj5014]

Thank you so much for conveying this topic in so much detail. I had been struggling with Time series forecasting for a while now. However, coming across your tutorials has been amazing and I am able to catch on very efficiently. God bless you sir

[@carrocesta]

Sir, you are a really good teacher. I am spaniard and I really understand everything you explain in here, very rare regarding neural networks. Thanks dude.

[@D.2023]

I am thrown into a bachelor project about times series forecasting, and all I have is introductionary ML course, mediocre descrete math and statistsics results with 1/2 intro-kalkulus course. Time-series forcasting is a major leap from 2-layered perceptron and linear regression, and some intro classification models... This stuff looks like to be my guiding light XD I actually understand what you say, and even if the math stuff kicks in - you throw in enough keypoint refference words for me to side-google the lacking understanding from somewhere else. So thrilled to discover your chanell, Thank You!

[@BharathSaiS]

This is the content I love to watch! Neat and crisp :)

[@lomash_irl]

This is a wonderful lecture, I found this relevant for climate data I was afraid all the data I have was white noise but applied statsmodels and found the decomposition helpful... For all those reading this comment, please suggest books or other relevant material I can read/learn to apply time series analysis on tropical cyclones Thank you Sreeni for being a great educator for us

[@kailashj2145]

Great explanation sir,thankful I found this channel

[@aqilahrose]

Omg, this is the great video I ever need. Would you mind if you can make it for multivariate time series forecasting using deep learning? (with lots of features or column).. look forward to this kind of topic from you!

[@nightline9868]

He dit exactly that

[@SmartPhone-tv4qi]

Awesome work , explained everything in single detail ,even why to use dark background! Look forward for more videos.

[@DigitalSreeni]

Thanks a lot 😊

[@tokiniainavictoriotrandria2028]

thank you for the Video Sir, it was exactly what i needed, you're amazing

[@TravelIsLove-22]

HI Sreeni, thanks a lot for the nice videos :) one question: in the step that you can the decomposed on the df, I get an error saying that I need to have a Datetime object with Frequency...it seems that I should assign frequency to df via df = df.asfreq('d'), how come that you didn't assign the frequency but the code works for you...

[@SAINIVEDH]

Thanks for the tutorial and How are you saying confidence intervals from the plot at 17:00 ?!!

[@Techog4Sale]

i spend my 3 days on arima but unable to get the main point, finally i come here and get the all relevant info....

[@aparnashrivastava5837]

Great video sir, Thank you so much

[@DigitalSreeni]

Glad you liked it

[@5112vivek]

Thanks Sreeni, Quick question, in ADF Test instead of looking at the p-value, for rejection, do we not look at the critical values. My understanding is that the asymptotics are different here, hence the critical values are provided by the ADF test separately. Thanks

[@aminabouhali5944]

Finally Ican understand easily

[@khairulfahim]

You are just awesome

[@МихаилСыч-л4с]

Лучший

[@iftikhar58]

thank you so much brother!!!!!

[@tchintchie]

great video as always. you could also very easily parse the dates when importing: pd.read_csv("/kaggle/input/air-passengers/AirPassengers.csv", parse_dates=["Month"]) if you know the date-column´s name

[@DigitalSreeni]

Thanks for the tip. Yes, once you feel comfortable with Pandas you can minimize the number of steps.

[@alteshaus3149]

Nice video sir

[@aminabouhali5944]

Thank you so much sir

[@ngyuentru8117]

What is lag? Is it just as its name suggests or is it a key term that means something else?

[@sumitmandal6676]

you can use " decomposed.plot() " instead of " plt.figure(figsize=(12,8)) plt.subplot(411) plt.plot(df['Passengers'], label='Original', color='yellow') plt.legend(loc='upper left') plt.subplot(412) plt.plot(trend, label='Trend', color='yellow') plt.legend(loc='upper left') plt.subplot(413) plt.plot(seasonal, label='Seasonal', color='yellow') plt.legend(loc='upper left') plt.subplot(414) plt.plot(residual, label='Residual', color='yellow') plt.legend(loc='upper left') plt.show() ". Though the one he taught is the best.

[@DigitalSreeni]

Thank you very much for the suggestion.

[@alamintalukder9950]

thanks a lot ❤

[@kotraner]

have to change adf, pvalue, usedlag_, nobs_, critical_values_, icbest_ = adfuller(df['#Passengers']) to prevent from invoking error.

[@u.fashion]

15:14