Handling Imbalanced Dataset in Machine Learning: Easy Explanation for Data Science Interviews
Вставка
- Опубліковано 24 чер 2024
- Imbalanced Data is one of the most common machine learning problems you’ll come across in data science interviews. In this video, I cover what an imbalanced dataset is, what disadvantages it presents, and how to deal with imbalanced data when data contains only 1% of the minority class.
🟢Get all my free data science interview resources
www.emmading.com/resources
🟡 Product Case Interview Cheatsheet www.emmading.com/product-case...
🟠 Statistics Interview Cheatsheet www.emmading.com/statistics-i...
🟣 Behavioral Interview Cheatsheet www.emmading.com/behavioral-i...
🔵 Data Science Resume Checklist www.emmading.com/data-science...
✅ We work with Experienced Data Scientists to help them land their next dream jobs. Apply now: www.emmading.com/coaching
// Comment
Got any questions? Something to add?
Write a comment below to chat.
// Let's connect on LinkedIn:
/ emmading001
====================
Contents of this video:
====================
00:00 Introduction
01:20 Interview Questions
01:38 Imbalanced Data
03:15 Why it causes problems?
04:27 How to deal with imbalanced data?
08:13 Model-level methods
11:33 Evaluation Metrics
13:25 Outro
Hi Emma, it is a really good summary videos on the matter of imbalanced dataset. Thank you and keep up the good work!
Thanks Emma , these short videos come in handy when preparing for interview
This video is amazing. It was easy to understand and summarized different possibilities for dealing with unbalanced data. Congratulations! Keep helping people. I am very grateful for your explanation!
I enjoyed this video. Thanks for this Emma
Great topic! Thanks for covering
Emma,great explanation and to the point.
This video helped me clear an interview. Subscribed. Thank you.
Best Video on ML, I understood very clearly. Thank You
This is really helpful. thank you so much for putting out these videos!
So glad you find them helpful, Daniel! Thanks for watching. 😊
Hi Emma,
these videos are really good.
can you make a video on time series analysis
Many of you have asked me to share my presentation notes, and now… I have them for you! Download all the PDFs of my Notion pages at www.emmading.com/get-all-my-free-resources. Enjoy!
is it possible to share your notion file? Thank you
@@jerrywang1550 You can download all the PDFs of my Notion pages at emmading.com/resources by navigating to the individual posts. Enjoy!
@@emma_ding I mean your notion files, not PDF. Thank you
Subscribed !!
Wonderfull!
Hi! Is there a way you can share this notion document! Thank you!! Great content
Thanks Emma, Can we also have a series of videos on deploying ML models in production?
Thanks for your comment, Sanyam! 😊 I've added your idea to my list of content suggestions.
To my view, imbalance of data does not pose a problem. During classification one ought to model class membership distributions, and these may be small. As long as they are correct, there is no problem. One should, of course, use proper scoring rules (i.e. not accuracy) to maximize the classification problem.
Tetlock's Superforecasting serves as a wonderful and very readable introduction to predicting unbalanced classes.
Checkout this paper on Gumbel loss/activation for LVIS long tailed dataset, interesting method for imbalanced datasets
Paper link?
?
Hey Emma..big fan of your work😀,looking for series in model deployment.. if you can add things like processing(batch/stream), serving(batch/realtime) and learning(offline/online) part in production. sorry if it is a big ask🥲
Thanks for your comment! I've added your suggestions to my list of content ideas. 😊
Hi Emma. Could you talk about chatGPT (including its model, dataset, algorithms, system design, etc) for the next video? Thank you.
Thanks for your comment! 😊 I've added your idea to my list of content suggestions.
Hi, Emma! Thanks for sharing. Very helpful materials. But i got a probleme when downloading the presentation notes, somehow the notes for imbalanced dataset is missing, when I click the imbalanced dataset notes, it actually opens the notes for encoding categorical data, could you please help with this?
Thank you so much for letting me know! I apologize for the mix-up, and have corrected the issue. Thanks for your patience. 💛
In the ‘why imbalance is important’ part, the accuracy for rare event predicting model can be solved by relying on other evaluating metric such as precision and recall, isn’t that right?. It’s not explaining the why
hey Emma please send me the code for imbalanced image datasets
You are just reading the text written in the book, try to explain with examples and further in detail, apart from what is already mentioned in the book.
Hi, audio clipping detected..
is 75:25 imbalanced dataset
A gorgeous ML scientist
please reply me
Your content is good, but your strong accent needs improvement.
So bad