Supervised vs. Unsupervised Learning
Vložit
- čas přidán 26. 07. 2022
- Learn more about WatsonX: ibm.biz/BdPuCJ
More about supervised & unsupervised learning → ibm.biz/Blog-Supervised-vs-Un...
Learn about IBM Watson Studio → ibm.biz/learn-watson-studio
Explore: IBM Cloud Pak for Data → ibm.biz/explore-pak-for-data
What's the best type of machine learning model for you - supervised or Unsupervised learning?
In this video, Martin Keen explains what the difference is between these 2 types, the pros and cons of each, and ... presents a 3rd possibility.
Get started for free on IBM Cloud → ibm.biz/ibm-cloud-tier
Subscribe to see more videos like this in the future → ibm.biz/subscribe-now
#AI #Software #ITModernization #IBM #MachineLearning #ml #watsonX
I love the way the content is delivered, very logical and clear. Thank you very much!
This instructor in particular has a wonderful way of explaining the topics very clearly and plainly.
Simple explanation, just what I needed.
IBM's content are the best I've seen
This is one of the best video on machine learning, short and precise 👍
I wish I can like this more than once
Much appreciated for such a useful tutorial video.
Thanks IBM.
Really helpful and quick. Thanks for the explanation!
Thank you Martin! Your explanation is great
Thank you for your explanation. Easier to understand than my uni coursebook.
So well explained, thank you!
It is way better than the university teachers .
U have the great skill to explain the subjects. Tnk u.
thanks for the wonderful subtitles. It helps me a lot to study this topic in English.
Nice clear description. Thank you!
This is an amazing explanation
Thank you, thank you, thank you for the clear explanations.
This is really an amazing video! Thank you so much! :-) really grateful for your help
Always amazing explanations
@5:29 official subtitles say "high risk given accurate results" while I'm 99% positive the speaker says "higher risk of inaccurate results". Huge divergence of meaning. Wish I knew how to ping the channel directly for clarification/correction.
Amazing explanation!! thank you
very nice video. great explanation and cool visual illustration
Very well explained
Thank you!
a much under-viewed excellent video
very useful and short to learn thankyou
thank you very much. It was very useful, helpful, clear and quick
Loved this! Thank you!
great video and great teacher
Excellent explanation!
Superb explaination!
it is very helpful.Thank you
I learn from you
Surprisingly 😊
wow..very impressive...thank you
good explaination
thank you brother!
amazing. thank you.
Great ! Thank u.
Thank you ❤️✋
Thanks a lot sir 💌
can you give more details about semi-supervised learning approach please ?? is this HITL ( human in the loop approach) ? whereas a small dataset is labelled where its being used to label and train other larger part of unlabelled dataset ??
Great video, very easy to understand! TY..
Any chance you can do a video on "Does Deep Learning suffer from Bias? If so, how? How can we overcome it?"
Thanks!!!!
How can he write in a speculare way so smoothly?
Search on "lightboard videos".
Machine Learning
Maybe he wrote regularily and they mirrored the video.
What does Speculate even mean ??? I don't know much english
Thx a lot man
Thanks
Thanks !👍🙏
You said Logistic Regression is for regression task but actually logistic Regression is a linear model for binary classification task .Thank you
Quite rightly so. Logistic regression outputs in binary which logically is a category/classification of true and false
Correct, logistic regression actually a classification model.
Well, can't you make predictions with association rules ?
If I make a synthetic copy of the original data (same feature distributions), alter the outcome with random numbers, and work on these artificial data to design an algorithm (i.e., define predictors, combinations of some where components are too rare, and range of values of hyper parameters for example) BEFORE I run the algorithm on the original, labelled data, can this first phase where I am blind to the outcome be named unsupervised learning?
wouldn't it be supervised learning as well? Since, it is labelled irrespective of how accurate your training data is, machine is going to learn from it.
I think this could also qualify to be a semi supervised scenario
great
I want to do climate change analysis what should I use ?
but why unsupervised data could not detect regression? If it shows out of average value, then that data might be a regression point. I am new to this and tried to use elk on my job, thanks
Regression in this context means predicting a continuous (non-discrete) value. It's not exactly the regression known is statistics.
classification sounds similar to clustering, what is the difference tho?
In classification, we know the number of available classes a priori (e.g. classifying tumors as either benign or malignant) and when training the model, we specify what class each training example belongs to.
On clustering, we may or may not know the number of clusters. Also, when training the algorithms we don't specify what cluster a training example belongs to (since we have no idea).
They are similar but not the same.
I want to know ho does this delivery of the video works …. I mean does he know to type mirrored or what is it … confused a lot
See ibm.biz/write-backwards
Thank you so much man …. Finally got answer to this 😅😅
I always wondered how it worked
Quick response tho 💯💯
Supervised = your dataset has a target variable
Unsupervised = your dataset does not have a target variable
5:00 you should have a video explaining what unlabeled data is. Here's a video I found helpful czcams.com/video/vSO8dFTtlfE/video.html
de-click your audio plz, i can hear your saliva
I'm a bit distracted by his ability to write backwards so easily and quickly.