Unsupervised Machine Learning

1 / 1

Unsupervised Learning






Slides


Please login to comment

31 Comments

Hi
If our data is labeled, we should use the supervisedalgorithm, but it can also be used forUnsupervised I want to compare which one works better?
Does supervised always work better than Unsupervised?

  Upvote    Share

Hi,

We generally use supervised learning to predict labels. On the other hand, we use unsupervised learning to find a pattern in the data such as outlier detection, reducing the dimensions, forming clusters, creating a hierarchy from the data, and many more.

So, we can't compare both as both are used for different purposes.

 1  Upvote    Share

Hi Team,

 

Is it possible to create a cluster for a kind of data which does not have numeric values. For instance

Let me know the steps to a cluster for such kind of data. 

Regards,

Birendra Singh

  Upvote    Share

Hello,

Can you please let me know how did you plot the new-points inside the cluster plots in Green? Pls. show me the code.

Thanks

  Upvote    Share

Hi,

That was just a visual representation and we do not have any code for that. However, it can be done using Matplotlib. You can check out our Matplotlib introductory project for more information:

https://cloudxlab.com/assessment/displayslide/5311/getting-started-with-matplotlib-step-1-introducing-matplotlib?playlist_id=480

Thanks.

  Upvote    Share

hi,

where can I find the CSV file used in starting to run on the local notebook?  I have downloaded the complete repository of the course from github, but unable to find points.csv in it, kindly help.

  Upvote    Share

Hi,

You can save the dataframe you created during the project using the following command, and then download them using the Jupyter dashboard to your local computer:

df.to_csv('file1.csv')

Thanks.

 1  Upvote    Share

Thanks

  Upvote    Share

Hello,

I'm not able to see the Unsupervised Learning notebook in ml/machine_learning, after a git pull also. Can you please help me

  Upvote    Share

Hello Team,

It might sound dumb but I have a question pertaining to supervised vs unsupervised techniques. I know I do not have any labelled data but if I can somehow label the data can I use dimensionailty reduction as preprocessing technique with Unsupervised learning to reduce the time for very large datasets. And how about PCA if I do not loose much info and only evaluate the main crucial compoents with unsupervised learning  by creating an instance of reduced features dataset..will labelling and unlabelling  part of unsupervised learning (in case of blind data for unsupervised learning) will make any difference.

Also , how clusterg is different from classification. I understood that Clustering is  based on moments which helps in finding AM and we  have centroids(like in mechanics) and on the other hand  classification essentially strives for a decesion boundray so its bit complex and time consuming as it  uses various optimization techniques to get at optimal solution based on user's choice of algo.

Regards,

HS

  Upvote    Share

Hi,

First of all, no question is "dumb".

Second, PCA itself is an unsupervised learning technique. However, if you label the data, the problem converts to a supervised learning problem.

And third, classification is used for supervised learning whereas clustering is used for unsupervised learning. The process of classifying the input instances based on their corresponding class labels is known as classification whereas grouping the instances based on their similarity without the help of class labels is known as clustering. As Classification have labels so there is need of training and testing dataset for verifying the model created but there is no need for training and testing dataset in clustering. Classification is more complex as compared to clustering as there are many levels in classification phase whereas only grouping is done in clustering.

Thanks.

  Upvote    Share

This was really helpful. Thanks for the detailed explanation now I can understand both in more depth.

Regards,

HS

  Upvote    Share

This comment has been removed.

Hi,

Please share the slides for Unsupervised Learning.

  Upvote    Share

Hi,

I have added the slides, please check if you can view them.

Thanks.

  Upvote    Share

Hi,

Now the slides are available.

Thanks.

  Upvote    Share

I am not able to open notebook as a result I am not able to submit answers, I have sent several mails with screenshots with no solution. I am really unhappy with cloudxLab service.

  Upvote    Share

Hi,

I am sorry to hear that. Let me help you with you issue. Have you tried restarting your server? Please follow all the steps from the below link and let us know if it works:
https://discuss.cloudxlab.c...
Thanks.

-- Rajtilak Bhattacharjee

  Upvote    Share

Hi, I have done everything , stop server, sign out, restart server and I have sent each and every screenshot, I am requesting you to check my mails sent today an hour back.

  Upvote    Share

Hi,

Could you please check once again and let me know if it is working fine.
Thanks.

-- Rajtilak Bhattacharjee

  Upvote    Share

Hi,
Now it is coming.
Thank you Rajtilak ????????????

  Upvote    Share

Where is the complete video, This video is just intro about clustering. @disqus_zQl19TrWvN:disqus when you are going to upload

  Upvote    Share

Please replace the video at the earliest to maintain the quality and standard of the course.

  Upvote    Share

Hi, Anil.

Thanks for your suggestion.
Many of the students had really appreciated the methods of teaching in this video as they were able to relate the concepts to their day-to-day things.

All the best.

  Upvote    Share

This lecture wasn't up to the standards set by the previous lectures, which were very well scripted and coherent. The first five minutes was all over the place. Felt very incoherent through out the lecture. Seemed a bit rushed and unprepared.

  Upvote    Share

This video is part of the live session. We will be soon releasing the scripted video for the same.
Thanks for the genuine feedback.

  Upvote    Share

when ?

  Upvote    Share

-- Please reply above this line --

Hi, Hemant.

This lecture is well explanatory. You can follow the tutorial steps by steps. You will be able to understand it.
Feel free to ask about the concept on ML, we will be able to happy to help.

All the best.
--
Best,
Satyajit Das

  Upvote    Share

Hi, Hemant.

This lecture is well explanatory. You can follow the tutorial steps by steps. You will be able to understand it.
Feel free to ask about the concept on ML, we will be able to happy to help.

All the best.

  Upvote    Share