Hands on Machine Learning - Chapter 2 - Full Machine Learning Project
Vložit
- čas přidán 2. 05. 2021
- PRACTICE DATA SCIENCE INTERVIEW Q's HERE: stratascratch.com/?via=shashank
A complete overview of Chapter 2 of the book Hands-on Machine Learning with Scikit-Learn Keras & Tensorflow
Dataset: drive.google.com/drive/folder...
You can get the book here: amzn.to/2SmaLBH
If you'd like to get the code along with much more soon to come please consider supporting me on my Patreon: / shashankkalanithi
FREE Python Tutorial: • Python for Data Analys...
Hardware:
MX Master 3 amzn.to/3sTroBW
LG 35in Curved Monitor amzn.to/39pPzR3
USB-C Hub amzn.to/31Ip8Sl
MacBook Pro Retina 16 Inch amzn.to/2PSwZde
Twitter: / kalamari95
I'm glad that at the end, you said that you did not want to withhold any knowledge, even when providing the code via Patreon. I respect that!
I am really happy to find your channel. The best video I watched so far in my data science learning journey. Thank you so much Sir.
As someone who is hesitant to start learning machine learning because I feel like "I'm not yet ready or well-versed enough in python", this is really, really easy to understand and wrap my head on. Thank you so much for taking the time and energy on putting this out for free. Cheers!
Oh my gosh!!! Hands on Machine Learning video too as well as Crab Statistics Book. Your channel is amazing.
I felt this book is so complex, thank you for making such videos
Please keep building this series
I have shared it to all my friends 🙌
I absolute love Sentdex and other high quality purveyors of programming knowledge on CZcams, however as a Machine learning beginner, this series in particular is really... REALLY well thought out- in both hands on coding but also logic / conceptual framework to wtf is actually going on. Thank you so much for putting this together, it is exceptionally well done.
Thank you SO much Roman!
Which playlist of Sentdex did you find more helpful
Love the content. You break down hard to digest stuff to what I can comprehend. Absolutely love it and appreciate explaining these concepts like I’m 5.
Thanks so much for the compliment! It’s what I’m trying to do
I really appreciate the thought put into this series, I hope you know just how grateful I am.
This deserves a lot more views! Man this is gold!!
This is great - hepful to go along with someone! Thanks for making this series!
Good work mate! Keep motivated!
Thanks for your effort on this. Its helpful for me who recently started with the book. Keep going! you are doing superb and looking forward to more
Damn your timing is impeccable. Just bought the book. Looking forward to the series !
Perfect! Let’s get through this together!
Woah was searching for something like this. I started the book this week. Ty!
Great. Absolutely subscribed and looking forward to this series.
Thank you so much man! Keep posted for more videos
This is quality content brother!
Thanks Marcus! Please check out my other videos as well
wow! great tutorial fam.. you're a wonderful teacher indeed..
More like this please, i like that it comes from a textbook and also simplified
Hi Shashank. I'm an Engineer from Spain trying to switch to this sector. Your content is amazing dude. Very well explained. And you just got a new Patreon subscriber just for providing the code for free. You are the kind of guy I would like to be friend of. Peace! (even though you speak quite fast I can understand your english lol)
Top tier video. Can't ask fore more.
You won my heart in the first few minutes......Insh'Allah will support your work brother. and Thanks keeping it free for everyone.😇
Thank you so much
.this is one of the most useful channels to learn ML. thank you
Love this.. Quality content 👍🏻
Great Content....Short and Clear
You are a good teacher. Thank you very much.
Thank you so much. Thank you VERY much!!!
Cool!! Go ahead!
Thank you for this amazing content
better than my online classes! Thanks
thanks Shashank. keep up the good work!
Thanks so much man!
Very interesting ! Cool ! TNX!
Thanks for checking it out!
Good work 👍👍 bro hats off to you!
Please continue this series.
Thanks man! New video coming out next week!
Thank you so much for the video
Excellent!! Greetings from Sweden!
Fantastic - looking forward to this!
You teach so gdamn well! Thank youu
Thank you Niko!
Thank you , so much.
0:00: 📘 The video series will cover the book 'Hands-On Machine Learning with Scikit-Learn and TensorFlow' chapter by chapter, simplifying the content for easier understanding.
6:39: ⚙ The video discusses the use of rules engine in predicting house prices and differentiates it from machine learning.
13:45: 💻 The video explains the importance of creating isolated environments for Python projects and demonstrates how to install scikit-learn using conda.
21:42: 📊 The video provides basic information on a dataset, including the number of entries, columns, and data types.
28:46: 📊 The video demonstrates how to create bins and labels for housing data and visualize it using a chart.
36:25: 💰 The importance of median income in determining housing prices and the impact on machine learning models.
44:22: ⚠ It is important to create a copy of the training data set to avoid overfitting to the test set and to maintain the integrity of the machine learning model evaluation.
51:45: 📊 The video discusses the strong correlation between median income and median house value, and the basics of data visualization using tools like Tableau or Power BI.
59:48: ⚙ The video discusses data transformation using the Imputer strategy in Python.
1:07:50: ⚙ Feature engineering involves combining and manipulating different columns to create better predictors for machine learning algorithms. Scaling the data is essential to ensure that all features are weighted appropriately.
1:15:37: ⚙ The video discusses creating a data frame to compare predicted house prices with actual prices using linear regression in machine learning.
1:22:49: ⚙ The video discusses fine-tuning a model by adjusting hyperparameters using grid search.
1:29:33: ⚙ The importance of visualizing data and preparing it for machine learning using feature engineering, imputation, and encoding categorical variables.
Recap by Tammy AI
Hi Shashank, thank you very much for the video, very useful. One thing I noticed, when you were predicting results with different models, you used training data set, which the model already knows. I think test data should've been used. Thanks again!
thank you!
When you pass strat_test_set to your function data_transformations, it appears as if the function is fitting both the OneHotEncoder and also the StandardScaler to the test dataset. If I'm not mistaken, we should only be transforming the test dataset based on the fit of our train data. Thoughts?
sir u helped me a lot sir plz make more videos like this u teach awesome thanks for making this video
Thanks much!
Thank you i was reading it just now
Love ur vids, looking forward to chapter 3!
Coming out in 30 mins :)
@@ShashankData nice! quick question, 51:40 you are talking about the "visible" strong correlation. in the graph the scale shows us the correlation is higher than 0.2 if its orange but that wouldn't be really high. only if we look into the table we would see the correlation is above 0.65. how would you adjust the scale from -1 to 1 to evaluate the corr better visibly? And it would be cool if you would move your cam in the recording to the bottom right corner where usually no coding is happening. would be great do see you type at 1:00:00 as example.
when I run the linear regression I get this error...any idea why?
ValueError: Found input variables with inconsistent numbers of samples: [16512, 4128]
Very good! makes the book! ~knuckle cracking at 1:19.17
Great video, very well explained each topic. It was easier to understand while going through the book along with the video. Even the 2nd chapter was long, very well explained.
I agree! The best way to watch these videos is with the books side by side
@@ShashankData yes I totally Agree
Thank you
very thankful if you continoue...
Love you bro
Thanks a lot for your content! It's really really helpful!
I have one question about the data, at Encoding Categorical variables: Is the attribute 'near the ocean' really not better than far from it? I mean, the houses near the ocean have higher prices than others, at least where I live. Thanks again for the help!
how did you handle the rows with null bedroom values?
Hi Shashank! I just came across your channel and I am enjoying the content you are creating. I am not sure if it is only me, but the volume on the video seems to be a bit low.
Could i use a different kind of scaler for this data?is it better to use minmax fpr asimmetric data?
Thanks
Hi Shashank really great work. Keep up. As I see the Chapter 1 is missing, and I was wondering if you maybe upload sometime in order to be a complete playlist? Thanks again Shashank
you are my angel man ... like suscribe and what you want ... this is high quality content that I'm looking for ... thanks a lot!!
I have that book in montreal, im in medellin now. Here i got data science from scratch, currenty on Python crash course
Nice! I have a Python crash course for free on my channel if you’re interested :)
Do you have a source code or resource for creating: One node task scheduler for reinforcement learning (RLScheduler) in cluster
appreciate the efforts thanks
Of course Vijaya! Thanks for watching
Thanks for sharing your knowledge with us .. What is the Documentation Tool you are using to document the process ? it looks so good . I also want to use it for documenting
Very useful Shashank. Please do more O'Reilly books.
PLEASE, here is the link for the CHAPITRE 1 ?? Thank you
Hi, how can I get a pdf version if I bought a book?
can you make a video explaining the performance measures and the difference between each
Great Videos!
I am working on a project for predicting a true/false output based on 3 input variables. What do you think the best ML method would be? Sorry if this is vague.
Regression family - supervised
hello shashank....i have found your videos very helpful and decided to take this initiative of learning python being a naive in this filed. however i setup the environment but now in VS code my file isnt read. After filepath command it doesnot show any path. It isn't opening like i can see in your video. Can you please let me know where am I lagging?
Nice video, keeping pic in the left side help to show more clear view
can i ask what i will be able to do after finishing this course ? or it is just kind of introduction?
Excellent sir Idris Italy
please keep going good content.. i subbed
Thank you so much Sharnjit! Let’s learn together!
At the right time....when I started reading this book 😀
Love it! Look out for chapter 3 next week!
Waiting for it ☺️😁
It's a great session, but u didn't mention the baseEstimator or describe the importance of it, i saw it in the book but i didn't geg what's the importance or usage of it. Could u explain it?
Form some reason the corr_matrix() at 49:39 gave me an error code 'could not convert string to float: 'NEAR BAY''. Do I have to drop the 'ocean_proximity' column? Great video btw,
I guess.. I was following the book before discovering this. It worked after dropping the "ocean proximity"
I couldn't find video of chapter 1
May I know what you're using to create the Mardown document (the one with the overview)?
Notion I believe
Kindly make more video like this. Waiting for chapter 3
Yessir! Coming out next week :)
i find your typing sound very satisfactory.
Thanks Fizip!
Hey man great videos !! Can you please create a playlist for all of your videos ..
Good idea!
Because we installed the packages in the specific environment, will we need to reinstall these packages for each environment we create in the future?
Yup. You need to install all the libraries every time you create a new virtual environment.
how many laptops do you have? Btw good content brah!
The Feature Engineer is different from the book, that is custom transform. can't use code
What happened to Lecture 1?
I am getting a value error in test labels
Can someone help me to solve??
me too, i have no idea why
(Edit: Found the problem, forgot to code the 'Concatenating with Categorical Variables' section
note sure why this throwing an error after i used strat_test_set on transformingData().
"ValueError: all the input array dimensions for the concatenation axis must match exactly, but along dimension 0, the array at index 0 has size 4128 and the array at index 1 has size 16512"
I don't know chat. I know shasank is busy and all. Anyone has had the same problem? Thank you btw.
I guess you are concatenating test and train data horizontally which is not possible as both have different dimensions. Check how you are concatenating
And one more thing brother.
This is a multivariate regression not univariate since system will use multiple features to make prediction not 1 feature
👌👌
In predicting a class at 09:58 … what is class in this context … if you give examples that would be helpful bro
Thanks
Bro please mention any python book that includes numpy, pandas
@Shashank Kalanithi Hello,
Please present the solution(visualizations) in another video or file.
Athough this part was very helpful but not complete. So please do it
PLEASE BRING CHAPTER 1 ALSO
The small screen which displays you typing is in the bottom left. Which makes it kinda harder to view your code.
It'd be much better if you shift it to upper right.
Good one 😍
But, sound is not enough
Thank you for all the effort however you escaped challenging topics like pipelines and transformers, it would be more beneficial to include them in the video.
Example 1-1. Training and running a linear model using Scikit-Learn can you please explain this from chapter one line by line?
Where's chapter 1? Or don't I have to go over chapter 1?
hey shashank, I must say that you have explained the chapter 2 very nicely,
I have one request like can you please share the link of that notion file aur can share the pdf of all these notes. like it will be very helpful for me.
Thank you...
Hey Akash thank you for liking the video! The notes are available on my Patreon
Oh My God i think i just found a GOLD !!