Azure Machine Learning Demo - Walmart Sales Forecast
ฝัง
- เผยแพร่เมื่อ 2 พ.ย. 2024
- Прогноз продаж сети магазинов Walmart используя исторические данные и регрессионный анализ Microsoft Azure Machine Learning Studio
Source Example and the Dataset gallery.azure....
Summary
Example of using regression model to predict sales based on Walmart historical data
Description
We will get data from kaggle.com and will predict sales for 45 Walmart stores located in different regions. Each store contains many departments, and we will predict sales for each department in each store. To add to the challenge, selected holiday markdown events are included in the dataset. These markdowns are known to affect sales, but it is challenging to predict which departments are affected and the extent of the impact.
Algorithms
In this experiment we will use Decision Forest Regression Algorithm for better prediction
Спасибо Вам огромное! Такого понятного и доходчивого объяснения еще не видел по этой теме. Детально и подробно, даже такому далекому от машинного обучения человеку как я стало понятно!
Пожалуйста!
Thank you for this amazing walk through!
You are welcome
I am getting following error when i followed same steps of the video-pls revert
Join Data Error
The size of "Comma-separated case-sensitive names of join key columns for L" is inconsistent with size of "Comma-separated case-sensitive names of join key columns for R". . ( Error 0027 )
If this problem persists, go to the forum for additional help.
I have the same erro, did you fix it??? Can you give me idea to fix it?
Я ещё добавлял день недели, а то получается много повторений одинаковых чисел, на практике это не так, суббота очень отличается от понедельника.
Да, вообще дату правильнее еще разбивать на день, месяц, год и добавлять еще день недели.
I've added English subtitles to video! Please use them for better understanding of training process
I see the scored values as same for most of the rows. Why is that so? Does it mean that for future weeks for different stores the sales is going to be same?
Hi, this just mean that or the features are not enough, or the store feature has less correlation than other features
@@DmitryKatson Does that mean we need to increase the features or more columns for this to come up with a proper prediction?
@@surbhishridhar4178 features. You can get them by adding new columns with new information or generate them
В каком случае String нужно переводить в категориальніе переменные?
в любом если String влияет на прогноз
Need source and dataset
gallery.azure.ai/Experiment/Walmart-Store-Sales-Forecasting-1
Very Good. May I know if you could do it in English,please? I have just watched the entire video and liked it very much.Thank you.
In English please!!!
Thanks Jeff for your comment! I've spent some time and added english subtitles. Hope this will help!
Yes, it does help a lot. Thank you soooo much.
i can't find this data :((((
Дмитрий подскажите пожалуйста сколько денег стоило провести этот прогноз на АzureML ? То есть сколько они берут за такие эксперименты/вычисления?
Владимир, это было бесплатно. У них есть бесплатный слой. Можете в прайс калькуляторе посмотреть. Только как будете создавать studio смотрите на web service app - и также проставьте бесплатный слой, там могут считать, или вообще web service app resource не создавайте (или удалите после создания)
hello i followed your steps till 6.19 of your video but i am getting store(2),store(3) duplicate values. Help pls!
same, did you fix it?
never mind I just used select column in dataset to remove the duplicates since the values are same after removal.
@@darichism yeah even if that columns exist they don't change any values
@@harunjoe8991 I had a same problem, do you have any idea to fix them? please help me!!!!!!
Thank you
I had a probem in 9:50, anyone had a same situation that can help me? Thank you