Time Series Analysis and Forecasting - Forecasting Sales in Python - ARIMA, AR, MA models theory

แชร์
ฝัง
  • เผยแพร่เมื่อ 21 ธ.ค. 2024

ความคิดเห็น • 77

  • @syahmiaiman2470
    @syahmiaiman2470 ปีที่แล้ว +1

    You absolutely have no idea how many abysmal videos I had to watch before I found yours. Thank you for your sharing

  • @maheshsale
    @maheshsale 4 ปีที่แล้ว +3

    Extremely nice video. The way you explain the things, is too good. You don't leave behind any doubt about the things that comes in each programming statement. Thank you very much.

  • @FinTaxPlanning
    @FinTaxPlanning 4 ปีที่แล้ว +4

    I have watched many youtube videos for time series but I really feel like this one helped me truly understand how to do it. thank you. My recommendation is that when you say " I covered this in my previous video" please tell us the name or title the videos in a way where we can follow a progression.
    Thank you so much for posting this.

  • @joel1234james12345
    @joel1234james12345 2 ปีที่แล้ว

    The way you explain is very simple and clear!! 💯

  • @makenayvonne244
    @makenayvonne244 3 ปีที่แล้ว +3

    The mean can be used to get an overall idea, or picture, of the data set. Mean is best used for a data set with numbers that are close together. The median can be used to get an idea of what values fall above the midpoint and what values fall below the midpoint.

  • @neelammishra5622
    @neelammishra5622 2 ปีที่แล้ว

    I really enjoyed watching your video. It clarified so many simple foundational concepts to me while learning through time series analysis. Thanks

  • @productvisibilitytvshoppin1107
    @productvisibilitytvshoppin1107 4 ปีที่แล้ว +1

    FANTASTIC! Well explained for the beginners like me to follow and understand. Thank You.

  • @fernandomaron
    @fernandomaron ปีที่แล้ว

    Amazing lecture! Many thanks 🙏

  • @amitgawade7168
    @amitgawade7168 4 ปีที่แล้ว +1

    Very insightful video sir...thank you!

  • @rudevariables
    @rudevariables 4 ปีที่แล้ว

    good video thx.. to answer one of your questions median is preferred when the data distribution is skewed and mean when its symmetrical

  • @emanali_
    @emanali_ 2 ปีที่แล้ว

    Very good course 👍 👌 😊

  • @tsehayenegash8394
    @tsehayenegash8394 ปีที่แล้ว

    I have 15 years temperature data, so how can I determine the trend by using multiple linear regression?

  • @kilaa0007
    @kilaa0007 3 ปีที่แล้ว

    can you help plz?
    i am getting this error when using the mean squared between the test and forecast :
    Found input variables with inconsistent numbers of samples

  • @rathnakumarv3956
    @rathnakumarv3956 2 ปีที่แล้ว

    how only the trend component , detrending is done?

  • @paroxymal7688
    @paroxymal7688 3 ปีที่แล้ว

    Sir, when I run the last cell, it takes 2 min to execute but no values of arima,tmse is being shown during or after execution, cell just executes that's all, pls help

  • @MsAaradhyaChopra
    @MsAaradhyaChopra 2 ปีที่แล้ว

    ValueError: Can only compute partial correlations for lags up to 50% of the sample size. The requested nlags 10 must be < 5.
    I am getting this error can you help in

  • @StanleyJohnCharles
    @StanleyJohnCharles 4 ปีที่แล้ว

    Hi.. When I give (3,2,3)..i am getting np. Sqrt(mean_squared_error(shampoo_test, shampoo_forecast)) as 75..can u check?

  • @jagdishbhapkar9739
    @jagdishbhapkar9739 3 ปีที่แล้ว

    very well explained thank you

  • @stonesupermaster
    @stonesupermaster 2 ปีที่แล้ว +1

    Hello Abhishek, thanks a lot for your video. I wanted to ask you if you've read how to apply forecasting models to time series with multiple SKU (like 500 - 2000) considering the efficienty while running it, thinking of using the forecast once every week. I would really appreciate if you can indicate me an study case or real case in which I can take a look to the approach within the code. Thanks in advance!!

    • @yohannanpchacko9896
      @yohannanpchacko9896 ปีที่แล้ว

      very important question, in real world we've to forecast multiple items dynamically, majority of the youtubers don't know this there is no such tutorials in you tube for multiple products forecasting using SARIMA and ARIMA

  • @abdullahsaif4312
    @abdullahsaif4312 3 ปีที่แล้ว

    Thank you a lot for this video but I have a question that comes to mind, why don't you take the sqr for the MSE in the iteration (for loop)?

  • @artificial-intelligence5753
    @artificial-intelligence5753 2 ปีที่แล้ว

    mean when data is symmetrical, meadian when data is skewed

  • @ravibengeri1507
    @ravibengeri1507 4 ปีที่แล้ว

    Hi Sir, What approach we should follow when the target variable is following sigmoid or logistic or S curve with respect to time.
    Shall we still apply Time Series? If we can which algorithm we should chose as it has multiple variables affecting target variable?

  • @anuragdeepak9975
    @anuragdeepak9975 4 ปีที่แล้ว +1

    Do we have to use stationary or original time series for training and testing purpose?

    • @Jam05_
      @Jam05_ 4 ปีที่แล้ว

      You can use both i guess!

  • @tejassutar4198
    @tejassutar4198 4 ปีที่แล้ว

    Sir if you plot that forecast & test data the graph not shows any relation?

  • @jitheshkarunakaran
    @jitheshkarunakaran 3 ปีที่แล้ว

    Hi Abhishek, I have items 100 their sales from Jan 2021 to Aug 2021. Can you explain how to predict the September sales of each item using Python ML?

    • @vipultyagi74
      @vipultyagi74 2 ปีที่แล้ว

      this is a classic case of multiple time series analysis. DeepAREstimator can be very helpful in this case

  • @shilpishankermishra
    @shilpishankermishra 3 ปีที่แล้ว

    Is it important to make timeseries stationary, to start working on MA? If any other assumptions has to be satisfied. Pls specify. TIA

  • @someshkumaryadav671
    @someshkumaryadav671 4 ปีที่แล้ว +1

    ValueError: Given a pandas object and the index does not contain dates , i am getting this error on executing the line df_model=ARIMA(df_train,order=(2,1,3)) in jupyter, sir can u please tell me how i can remove it

    • @AbhishekAgarrwal
      @AbhishekAgarrwal  4 ปีที่แล้ว +1

      To me it looks like dates are not properly parsed and that's why it is giving error message "index doesn't contain dates". While reading csv, there is a parameter prase_date parameter.. try using that on date column.

    • @someshkumaryadav671
      @someshkumaryadav671 4 ปีที่แล้ว

      @@AbhishekAgarrwal thanku for replying but i had solve this problem just on next line another error is coming that.. "typeerror: cannot cast ufunc subtract output from dtype('float64') to dtype('int64') with casting rule 'same_kind') help please...

    • @RVideoTutorials
      @RVideoTutorials 4 ปีที่แล้ว +1

      @@someshkumaryadav671 What is the code written in that line?

    • @someshkumaryadav671
      @someshkumaryadav671 4 ปีที่แล้ว

      @@RVideoTutorials model_fit= model.fit() after executing the ARIMA

  • @theabhishekkumardotcom
    @theabhishekkumardotcom 4 ปีที่แล้ว

    nicely explained, could you also extend this to explain how to check/ test for various components presents in a time series data, and add more videos to build upon the modelling to find best time series forecasting for the shampoo data. Thanks in advance 👍

  • @puspanjalisarma4711
    @puspanjalisarma4711 3 ปีที่แล้ว

    Hi Abhishek, How do I build an ARIMA model with multiindex(2 columns taken in index)? I am getting error and not able to move forward.

    • @SaurabhYadav-fs7si
      @SaurabhYadav-fs7si 3 ปีที่แล้ว

      no you can't have 2 dependent variables because time series is a supervised approach..

  • @girishlc
    @girishlc 4 ปีที่แล้ว

    Hi Abhishek, thank you very much for the clear explanation,
    Can you please provide the link or explain the python implementation for "Multivariate Time series" problem.

  • @earthexplorer579
    @earthexplorer579 4 ปีที่แล้ว

    Hi i need make a forecasting program for a project, do you write customize programs to sell ?

    • @AbhishekAgarrwal
      @AbhishekAgarrwal  4 ปีที่แล้ว

      Yeah.. you can share the requirements with me @ abhishek.agarrwal@gmail.com and I can evaluate and reply.

  • @tech9ine904
    @tech9ine904 4 ปีที่แล้ว

    Hi, thanks for nice explanation but there is a problem, when you calculate rmse for the order 312 it is 130.92, but for same order when you calculate the rmse through looping different pdq values you got 134.51 for same order 312. How is that possible? i tried this code with some other datasets and i came across same errors.

  • @gaurisingh3198
    @gaurisingh3198 5 ปีที่แล้ว

    where u have calculated the forecast of upcoming years

  • @mechdeeps
    @mechdeeps 4 ปีที่แล้ว

    Thank you Abhishek for explaining in such a detail.
    If we finalize the pdq values from permutation and combinations, when do we have to consider pdq from Pacf and Acf plots?

    • @terrellantonio5319
      @terrellantonio5319 3 ปีที่แล้ว

      i realize it is pretty randomly asking but does anyone know of a good website to watch newly released tv shows online?

    • @waylonjoshua5273
      @waylonjoshua5273 3 ปีที่แล้ว

      @Terrell Antonio flixportal :)

    • @terrellantonio5319
      @terrellantonio5319 3 ปีที่แล้ว

      @Waylon Joshua thanks, I signed up and it seems to work :D I really appreciate it !!

    • @waylonjoshua5273
      @waylonjoshua5273 3 ปีที่แล้ว

      @Terrell Antonio glad I could help :D

    • @beyourself2209
      @beyourself2209 3 ปีที่แล้ว

      U can use acf and pacf plots to predict q and p. But there might be issue with exact parameters selection. Hence it is better to use grid search to get the fine tuned parameters .

  • @ashishdiwe4110
    @ashishdiwe4110 3 ปีที่แล้ว

    hii sir,
    I'm trying to build a time series forecast model for my final year project wherein I can predict the sales and demand for hand sanitizer. can you please help me out.

    • @vipultyagi74
      @vipultyagi74 2 ปีที่แล้ว

      yes I can help. Give me more details and any particular challenge you are facing

  • @ashZay
    @ashZay 4 ปีที่แล้ว

    your model and teaching technique is awesome. I tried running model but in loop syntax, it showing me a syntax error. I was practicing on real data. Can you please help ( for p in p_values .....)

  • @helpyburhanuddin3087
    @helpyburhanuddin3087 4 ปีที่แล้ว

    Excellent.....kindly show me how to work with yearly data please instead of months.....

    • @vipultyagi74
      @vipultyagi74 2 ปีที่แล้ว

      the method is not gonna change. be it any timestamp(yearly, monthly, weekly, daily) provided timestamps are equidistant from each other.

  • @fobembe
    @fobembe 4 ปีที่แล้ว +1

    great job. I wanted to purchase the notebook but could not succeed. You dont have paypal?

    • @AbhishekAgarrwal
      @AbhishekAgarrwal  4 ปีที่แล้ว

      yes.. you can pay to the abhishek.agarrwal on paypal and then send me the email on abhishek.agarrwal@gmail.com

  • @debatradas1597
    @debatradas1597 3 ปีที่แล้ว

    Thank you

  • @littlecreator4838
    @littlecreator4838 2 ปีที่แล้ว

    Sir, I am eager to learn from you. Pls let me know if you provide online trainings.

  • @mudassirqadeer2782
    @mudassirqadeer2782 4 ปีที่แล้ว

    good explanation ; @6:10 "Kepted" lol

  • @sadafwaqas5972
    @sadafwaqas5972 4 ปีที่แล้ว

    cannot convert data frame into time series? why is that?

  • @nikhiljagtap6799
    @nikhiljagtap6799 4 ปีที่แล้ว

    Sir mera arima model (0,0,0) aya hai to mai kya karu ?

    • @AbhishekAgarrwal
      @AbhishekAgarrwal  4 ปีที่แล้ว

      You can check if you are getting best results while testing this on test data.

  • @avinashshanmonteiro6075
    @avinashshanmonteiro6075 4 ปีที่แล้ว

    You didn't show how to make the model stationary.

  • @CodeWithMuh
    @CodeWithMuh 4 ปีที่แล้ว

    please share github link of this code .

  • @Emotekofficial
    @Emotekofficial 2 ปีที่แล้ว

    If there are Outliers you use Median as central tendency, else Mean!

  • @RidingWithGerdas
    @RidingWithGerdas 5 ปีที่แล้ว +5

    Another tutorial where nobody shows how to predict out of sample values which would not be mean..

  • @teenamertia5704
    @teenamertia5704 3 ปีที่แล้ว

    cointegration please

  • @dataarchitect
    @dataarchitect 2 ปีที่แล้ว

    You didn't explain the ARIMA order properly.

  • @kannadastocktrader3369
    @kannadastocktrader3369 ปีที่แล้ว

    dont u think what u explained is 1/2 knowledge??? i didn't like the way u approach