Access Hindi, English courses here- www.unfolddatascience.com/s/store Plz register on the website
@GopiKumar-ny3xx
4 жыл бұрын
Useful information... Nice presentation
@UnfoldDataScience
4 жыл бұрын
Thanks a lot :)
@piotr3024
3 жыл бұрын
You are amazing, thank you for being so supportive and explaining this with ease
@UnfoldDataScience
3 жыл бұрын
Thanks Piotr.
@mrsmelaniecook
Жыл бұрын
Thank you very much! You are a life saver!
@rodrigoarrudacarneiro8457
4 жыл бұрын
Thank you very much for the excelent video. It was all very clear. I've noticed you open your drive on google for codes downloading and i just did for studies purpouse. Thank you very much for the excelent job
@UnfoldDataScience
4 жыл бұрын
Thanks you, please find my codes and data used here: drive.google.com/drive/folders/1XdPbyAc9iWml0fPPNX91Yq3BRwkZAG2M
@anugrahmuzakkipuar5020
2 жыл бұрын
Thank you for your video! I read some theory about this topic but this is the first time I see it in action directly!
@UnfoldDataScience
2 жыл бұрын
Glad it was helpful!
@shwethajaison8740
2 жыл бұрын
Awesome
@UnfoldDataScience
2 жыл бұрын
Thank you
@arjundev4908
4 жыл бұрын
This video very helpful was asked in interview.. Thank you very much!
@UnfoldDataScience
4 жыл бұрын
Glad it was helpful Arjun. happy Learning! Stay safe.tc
@austindsouza3145
2 жыл бұрын
Good content
@marcymierwaldt2727
3 жыл бұрын
Thank you so much for this helpful and very well explained content! I am so grateful I found this video. It was great to support my university seminar work!
@UnfoldDataScience
3 жыл бұрын
Glad it was helpful!
@ferozsharif6281
2 жыл бұрын
Comprehensively explained the concept, thank you.
@UnfoldDataScience
2 жыл бұрын
You're very welcome!
@RameshYadav-fx5vn
4 жыл бұрын
Very nice
@UnfoldDataScience
4 жыл бұрын
Thank you :)
@sadhnarai8757
4 жыл бұрын
Very nice Aman
@UnfoldDataScience
4 жыл бұрын
Thank you :)
@YashpalNSharma
4 жыл бұрын
Hi Aman. Great video again. Couple of questions - 1. How can we determine the acceptable/threshold values for Support, Confidence and Lift? 2. In Recommendation Systems, are Antecedent and Consequent reversible? Can they be reversely recommended?
@UnfoldDataScience
4 жыл бұрын
1. We can not determine in advance. By looking at the output, in the next iteration we can play around with threshold. 2.It makes sense to me.
@codeloki
2 жыл бұрын
Yes, but you see the problem here, The assumption is product bought by maximum people is what new customers will also buy, 1.Possibility in such recommendations, you will end up giving the aggregated labels all the time, no personalization to the incoming customer, maybe dig in customer segmentation first. rank the products vs customer and then predict, or if they all follow the same plane may be SVD can help. After that put rule mining to see the aggregation shift in the customers buying behavior in each season. And prepare your model for next season with max strength or max likelihood of a particular customer. Hope it'll help.
@ashaikh147
6 ай бұрын
easy to understand. really nice one
@mohammadsirajulislam6290
Жыл бұрын
Your job is nice. I do this earlier, but i enjoyed with you.
@AI-NotesExperience
10 ай бұрын
Well done and best wishes
@UnfoldDataScience
10 ай бұрын
Thank you
@nickwu5317
4 жыл бұрын
Greate content, looking forward to more demos.
@UnfoldDataScience
4 жыл бұрын
Thank you, happy learning. Tc
@siddhantjain452
2 жыл бұрын
Ek no video !!
@UnfoldDataScience
2 жыл бұрын
Thank you
@sandipansarkar9211
2 жыл бұрын
finished watching
@UnfoldDataScience
2 жыл бұрын
THanks
@the_senthil
3 жыл бұрын
Worth!
@UnfoldDataScience
3 жыл бұрын
Thanks a lot.
@dev5289
3 жыл бұрын
Great video
@UnfoldDataScience
3 жыл бұрын
Thanks for the visit Devansh.
@venukumarvenk
3 жыл бұрын
3:58 not credit transactions, they are cancelled transactions, as you will see negative values in quantity for those records
@UnfoldDataScience
2 жыл бұрын
Ok thanks for pointing out
@jayminsinhbarad4335
Жыл бұрын
Can algorithm program are place upto 10 baskets order at one time under 1 sec?
@nandnigayakwaar6193
4 жыл бұрын
Nice explanation, sir
@UnfoldDataScience
4 жыл бұрын
Thanks Nandni. Happy learning. Tc
@dinosilooy2228
4 жыл бұрын
Thank you so much for this useful tutorial!
@UnfoldDataScience
4 жыл бұрын
You're very welcome!
@rajat1548
4 жыл бұрын
Great content
@UnfoldDataScience
4 жыл бұрын
Thanks Rajat. Keep Watching.
@UnfoldDataScience
4 жыл бұрын
Thanks Rajat. Keep Watching.
@АнжеликаСемынина
4 жыл бұрын
So great lesson! Thank you :)
@UnfoldDataScience
4 жыл бұрын
Most Welcome. Thanks for watching :)
@dataframe2635
4 жыл бұрын
i mean why . ye jisne bhi unlike kiya he video... bhai..kya problem he... kya acha nai laga vo likh na comment me. itna acha video he firvi unlike.
@UnfoldDataScience
3 жыл бұрын
Thank you :)
@rohanshet123
2 жыл бұрын
thanks
@UnfoldDataScience
2 жыл бұрын
Welcome Rohan
@animikhchakraborty8224
4 жыл бұрын
Good explanation. Why we clean the spaces in data preparation ? Please say something about data preparation. This part remains unclear.
@UnfoldDataScience
4 жыл бұрын
There can be many data preparation steps. The main thing to take care in this case is cleaning junk in the data, and preparing data in transnational format.
@P4uluv
4 жыл бұрын
Nice video and explanation. I'm using a different dataset. While it gives me the frequent itemsets, it only displays the heading when I'm trying to get the rules even though I have reduced the threshold to 0. Any suggestion?
@UnfoldDataScience
4 жыл бұрын
Hi Muhammad, Try to lower down support, lift, confidence all. Do run without any values to see what are the lowest values coming.
@P4uluv
4 жыл бұрын
Thanks. It's now working for some reason.
@goutamborthakur4980
4 жыл бұрын
Awesome lesson! Thanks Aman! Could you please help me out with the below points: 1. How can we say that which support, confidence and lift values are good? 2. How can we consider lift >= 3 and confidence >=0.3 values?
@UnfoldDataScience
4 жыл бұрын
Hi Gotham, First answer - Cannot fix a number here, depends on how many rules model is giving on default values and then we can change thresholds to get more/less "strong rules". Second answer - When you call the apriori algorithm in R/python you can pass your threshold as parameters itself.
@goutamborthakur4980
4 жыл бұрын
@@UnfoldDataScience , thank you very much for your reply!
@irfankhanhmt
Жыл бұрын
❤
@UnfoldDataScience
Жыл бұрын
Thanks Irfan
@hpsa-stepstep479
2 жыл бұрын
Hello, thank you for shearing us your knowledges. Now I have dataset which have more than 300gb, when i try to run this algoritm the kernel shut down, can you recommend me another algoritm or way to work market basket analysis with this huge dataset?
@gitanjalikumari3749
3 ай бұрын
out of 1695 column how do we know which column is nor required ? as you mentioned postage column is not required and dropped it, how will we know the same ? please explain
@jaswanth174
4 жыл бұрын
Hi sir!I liked ur explanation ,its amazing .Can u explain using Fp growth algorithm?
@UnfoldDataScience
4 жыл бұрын
Hi Jaswanth, yes will do. Thanks for the ask. happy learning!
@jaswanth174
4 жыл бұрын
Sir...many people have asked about fp growth algorithm...it's been a long time I've asked!Please upload it as soon as possible sir...it's really urgent!
@susmithakesanam3547
4 жыл бұрын
Sir can you please explain using fp growth algorithm for the same dataset..I just need as simply as u explained in this video..please sir
@UnfoldDataScience
4 жыл бұрын
Hi Susmitha, yes I will create once I am done with ongoing playlist on ANN and deep learning.
@susmithakesanam3547
4 жыл бұрын
Ok sir thanku😇 May I know when will u do that?
@kanchidoshi6907
2 жыл бұрын
Can you please share the link of the videos for the explanation of lift confidence and support concepts?
@chandurockz8148
4 жыл бұрын
Hi sir!Nice explanation.I would like to know whether u can create using Fp growth algorithm or not?
@UnfoldDataScience
4 жыл бұрын
Thanks Chandu, yes using fp i will explain as well. happy lerning.tc
@akshay86523
3 жыл бұрын
Hello Aman, thank u for great video. Loved it and very informative. I have a question, while creating the basket you selected only germany.. what if I have to do it for the entire dataset (For all countries at once). how do we do it? I tried it but I'm getting an error saying grouper and axis must be same length. please advice? Thank you
@UnfoldDataScience
3 жыл бұрын
hi Akshay, this seems to be a format issue. See this link once: stackoverflow.com/questions/19483991/how-to-make-grouper-and-axis-the-same-length
@sampadathorat5885
3 жыл бұрын
How to validate trained model over testing datase?? How can these rules be applied to predict values for test datase?
@UnfoldDataScience
3 жыл бұрын
HI Sampada, the rules that are generated can be formatted and used as rule against new product.
@pratibhashinde7786
2 жыл бұрын
What if I'm the first customer? Then what will recommend?
@joelt83
3 жыл бұрын
Could you please do this in 'R'?
@UnfoldDataScience
2 жыл бұрын
This can be done using aproiri package in R
@sanduniprarthana1416
3 жыл бұрын
Thank you so much...could you please explain what is that column 'POSTAGE'?
@UnfoldDataScience
3 жыл бұрын
Let me check.
@sanduniprarthana1416
3 жыл бұрын
@@UnfoldDataScience please can you explain answer for my question?
@farhanmalhotra483
3 жыл бұрын
@@sanduniprarthana1416 have the same doubt
@farhanmalhotra483
3 жыл бұрын
@@UnfoldDataScience could you please clarify why the 'POSTAGE' column was dropped!
@lovelyjain4628
3 жыл бұрын
Caan u plz tell how to recommend now? like input will be product A, then recommend products related to it? how to do this part?
@UnfoldDataScience
3 жыл бұрын
Yes, for a rule "bread" => "milk" with high lift support confidence, if user buys "bread" then "milk" should be recommended.
@lovelyjain4628
3 жыл бұрын
@@UnfoldDataScience yes, can u share code for this...how to get this output for input milk...
@yohannesayana9456
2 жыл бұрын
Can we do both apriori and fp growth algorithms together?
@sarfarazansari6293
4 жыл бұрын
Hi thank you for the video! I wish to know the code if we have yes and no or 1 and 2 kinda categorical data how to factorize it as In Arule we have convert cat data into factor. Which we do in R as. Factor
@UnfoldDataScience
4 жыл бұрын
Hello Sarfaraz, I am not sure if I got your question. Do you mean "yes" or "no" as items? or target? its not a supervised learning just to mention.
@mind-your-body
4 жыл бұрын
Sir we can do by pivot table also data mining after that apply the apriori
@UnfoldDataScience
4 жыл бұрын
Yes we can.
@mind-your-body
4 жыл бұрын
Thank you sir for replying
@joebasshd
Жыл бұрын
How do you check to see the highest association?
@UnfoldDataScience
Жыл бұрын
Sort by values.
@ulisesprev91
4 жыл бұрын
Thank you so much for this video Aman, it was very helpful for me! I did the same analysis with my own data frame but I am running through a problem hope you can help me out. I have around 15,000 different SKUs therefore the Support for each product is pretty low, hence the lift. So I am only getting like even with a very low value of support (around 0.007) 15 rules and most of them are for the same products. When I try to reduce the support constrain to less than .0055 I got the blue screen on my computer. Any suggestions on how to do the analysis for a lot of SKUs? and another question, is it a way to make rules for each and every product? or at least the vast majority?
@UnfoldDataScience
4 жыл бұрын
If there are not much variation in the baskets then you will get only few regular products in many rules. Try to get more data which has variation inside the basket. To answer the blue screen problem, it might due to processing issues since you are bringing threshold to very low value, as a results many calculations are going inside hence system is hanging. Even if there are many SKU but no variation, you might not get many useful rules.
@ulisesprev91
4 жыл бұрын
Unfold Data Science thanks! I’ll work on that!
@yogeshsubramanyam2132
4 жыл бұрын
Sir thanks for uploading the video, If I perform Basket Analysis for UK do I need to tweak the code.
@UnfoldDataScience
4 жыл бұрын
Hi Yogesh, No same approach is supposed to work, just the input data should be in transaction format.
@yogeshsubramanyam2132
4 жыл бұрын
Sir I have tried but it has no error while executing but the output does not displaying the respective values of antecedent,consequent,suport etc
@tejateja8474
4 жыл бұрын
Sir could you please explain FP growth algorithm!
@UnfoldDataScience
4 жыл бұрын
Hi teja, Noted. Will create video on it.
@laibaasghar9771
6 ай бұрын
Sir where u can download this data set
@DS_AIML
4 жыл бұрын
@Aman,How can we recommend top 3 products to EACH user based on their previous purchase history with same data set using KNN ?
@UnfoldDataScience
4 жыл бұрын
Hi Anjani, the first approach can be "user based collaborative filtering". I have created detailed playlist on recommendation engines. You can have a look and also KNN can be one of the approach. I will put a video on this topic as well. Thank you Aman
@bhargavsolanki6386
3 жыл бұрын
Is it apriori algorithm?
@UnfoldDataScience
3 жыл бұрын
Yes Bhargav.
@creator025
4 жыл бұрын
Is apriori ML algorithm..?
@UnfoldDataScience
4 жыл бұрын
Yes Sumit.
@deekshanayak6572
9 күн бұрын
Dataset?
@ivanariyanto4412
3 жыл бұрын
how do you define the value association rules of that data sir ?
@UnfoldDataScience
2 жыл бұрын
Value as in? I may not be getting question
@livelovelaugh4050
3 жыл бұрын
Generally we load .csv or excel file for learning purpose . But how the data is loaded in real project ? Can you please help .
@UnfoldDataScience
3 жыл бұрын
That is what I am covering in my Unix deployment videos, watch latest videos on deployment Geeta.
@livelovelaugh4050
3 жыл бұрын
@@UnfoldDataScience Thank you 👍
@keerthanashankar7530
3 жыл бұрын
How to install Mlxtend package ... please tell me bcz I had error
@UnfoldDataScience
3 жыл бұрын
Hi Keertha, What error you are getting, I was facing some issues but I was able to get some help online and then install.
@nuwanperera7744
Жыл бұрын
Hi Aman, Please send me a link I can get Retail data sets for my degree project.
@mujahidislamkhan782
4 жыл бұрын
can you please explain me how to do this using sklearn/scikit library
@UnfoldDataScience
4 жыл бұрын
This package is not part of sklearn or scikit as of now.
@bangtansonoli1797
Жыл бұрын
Can you please provide us this dataset?
@chintalasivaniyit
3 ай бұрын
dataset?
@nandnigayakwaar6193
4 жыл бұрын
What is anticedeant support and consequent support
@UnfoldDataScience
4 жыл бұрын
{item1,item2} => {item3,item4} If above is a rule. then Here {item1,item2} are called anticedeant and {item3,item4} are called consequent Definition of support, you must be aware. That is how those terms are defined. Happy Learning! Thank you Aman
@nandnigayakwaar6193
4 жыл бұрын
@@UnfoldDataScience thanks
@shravanikummari9574
4 жыл бұрын
hi getting error while importing mlextend and apriori. any inputs? File "C:\Users\SC-ASS-\Anaconda2\lib\site-packages\mlxtend\frequent_patterns\apriori.py", line 47 yield from old_tuple ^ SyntaxError: invalid syntax
@shravanikummari9574
4 жыл бұрын
Got it. My python was 2. Updatedit to 3
@UnfoldDataScience
4 жыл бұрын
ok :)
@rkreddy7707
4 жыл бұрын
Hi sir..actually I'm doing a project on market basket analysis but using fp growth algorithm.I've tried a lot but I couldn't make it out.Could u please help me as soon as possible sir!
@UnfoldDataScience
4 жыл бұрын
Yes. FP is asked by other as well. I will create.
@apoorvaphadnis4741
4 жыл бұрын
After running "my_rules = association_rules(my_frequemt_itemsets,metric="lift",min_threshold=1)", I am getting error as AttributeError: 'generator' object has no attribute 'columns' What is expected here?
@apoorvaphadnis4741
4 жыл бұрын
I have read my data in csv format. Should I be converting it to transactions?
@UnfoldDataScience
4 жыл бұрын
Hi Apoorva, yes the algorithm expects you to supply transactions as input.
@apoorvaphadnis4741
4 жыл бұрын
@@UnfoldDataScience i tried searching for converting the read data into transaction in python but could only find for R. Also, I didn't find anything about conversion in the video? Could you please guide (syntax)
@shivanathmahalingam3466
4 жыл бұрын
If we run for all countries, there is a memory problem while loading apriori. How should we overcome that? Thanks
@UnfoldDataScience
4 жыл бұрын
Hi Shivanath, please limit number of rows that goes as input. Also you can change the support/lift/confidence numbers so that less records pass the criteria. Thank you Aman
@ahmedserag2245
4 жыл бұрын
what is version payton are y used
@UnfoldDataScience
4 жыл бұрын
Hi Ahmed, I am using 3.7
@rupenchitroda2471
4 жыл бұрын
Hey Aman! I am getting "AttributeError: '_csv.reader' object has no attribute 'size'" error. Can u please explain?
@UnfoldDataScience
4 жыл бұрын
Hi Rupen. This seems to me a python version issue. Do you have Python 2 or 3?
@rupenchitroda2471
4 жыл бұрын
@@UnfoldDataScience python 3.7
@vishwaskumar8604
Жыл бұрын
where is dataset how i will download
@trippics8465
3 жыл бұрын
don't understand why didn't you run #converting all positive values to 1 and everything else to 0?
@UnfoldDataScience
3 жыл бұрын
Thanks for suggesting, Let me review once based on your suggestion.
@varunthakur8545
4 жыл бұрын
Sir, you have uploaded a different dataset in the Market Basket folder in Google Drive , MBA.csv .. Can u tell how to work with that dataset?
@UnfoldDataScience
4 жыл бұрын
Hi Varun, i think it is the same file. If not, please convert it to type transactions and proceed.
@tinakovacova4412
3 жыл бұрын
Here is the link to correct dataset: archive.ics.uci.edu/ml/datasets/online+retail
@abuyudhistira4129
3 жыл бұрын
Hello sir, actually i am working on such a project similar with it, but the rows i have reaching almost 17 millions and impossible to use pandas. Hence i use spark with pyspark, is it possible to run the library on spark environment? If you could provide the codes also would be very helpful, thanks sir
@UnfoldDataScience
3 жыл бұрын
HI Abu, you need to find the equivalent is spark. I am not very sure may be I will do some research and come back. Your problem is very genuine.
@princeeaso7457
3 жыл бұрын
sir but the dataset u shared in drive is not working here.
@princeeaso7457
3 жыл бұрын
sir pls reply how can i find that data set
@UnfoldDataScience
3 жыл бұрын
Hi Prince, even if does not work with my data, you can google this topic, find some good links and use the same data they hv used, concept remains same.
@princeeaso7457
3 жыл бұрын
@@UnfoldDataScience thnk u sir i got the dataswt from the link u posted on the notebook. U are amazing
@UnfoldDataScience
3 жыл бұрын
Welcome Prince.
@ReelVibesGram
11 ай бұрын
sir dataset link
@sergiocoutinho6133
4 жыл бұрын
Hi, is it possible to donwload python code and execute it locally? Best regards
@sergiocoutinho6133
4 жыл бұрын
I am sorry .. i have found it at : drive.google.com/drive/folder... .. regards
@UnfoldDataScience
4 жыл бұрын
Thanks for watching. Stay Safe. Take care :)
@susmithakesanam3547
4 жыл бұрын
Sir! I've been waiting since 1month sir... U r not uploading video on FP growth....at least u r not replying
@UnfoldDataScience
4 жыл бұрын
Will upload soon for sure.
@freestyle2432
4 жыл бұрын
*** how to integrate this code in our Ecommerce site. Can you please do this ? ***
@UnfoldDataScience
4 жыл бұрын
Can be done, more data is needed.
@freestyle2432
4 жыл бұрын
Can you make short video regarding this? Just a complete process cycle?
@kidzaniacartoon7830
4 жыл бұрын
Sir, could you please share the dataset
@UnfoldDataScience
4 жыл бұрын
Hi Soma, I have uploaded the data file in my google drive. link in description
@kidzaniacartoon7830
4 жыл бұрын
@@UnfoldDataScience Thank you so much Sir, its a great help
@erumashraf2811
4 жыл бұрын
I need fp-growth algorithm code in python. can someone help please?
@UnfoldDataScience
4 жыл бұрын
Hi Erum, I found a link that might be helpful. Check this out: towardsdatascience.com/understand-and-build-fp-growth-algorithm-in-python-d8b989bab342 Stay Safe. Happy Learning!
@erumashraf2811
4 жыл бұрын
@@UnfoldDataScience Thx for your effort. This code is counting the number of input values. I have to count the frequency of each "input value" in the given dataset and then applying all the steps required for fp-growth to produce association rules. I have found code in a youtube link kzitem.info/news/bejne/xISht4RuZomhqGk but that is not completed.
Thanks for this video. I have used the same code on a very similar data structure but the code block def encode_units(x): if x = 1: return 1 is throwing up an error for me. TypeError: '
@UnfoldDataScience
3 жыл бұрын
Try to call function like this: encode_units(int(yourvalue))
@mind-your-body
4 жыл бұрын
Sir the implementation is same
@UnfoldDataScience
4 жыл бұрын
Did not get ur question Hemang.
@mind-your-body
4 жыл бұрын
I means this implementation remains same for all the apriori algorithm ? My question was that
@UnfoldDataScience
4 жыл бұрын
The way you pass data and you get output remains same.
@sauravdas6416
4 жыл бұрын
Please give the github code link
@UnfoldDataScience
4 жыл бұрын
Hi Saurav, you can find all codes in my google drive. Link in Description. Happy Learning.
Пікірлер: 189