Summary: 1. Choose a good project 2. Experiment X 3 3. Write good & understandable code 4. Share your code with good documentation (ReadME) 5. Mention License 6. Package your project 7. Make a good looking web application 8. Make a cool demo 9. Write an article/blog 10. Share it on linkedin/twitter etc
i'm currently reading your book. I LOVE IT. I take a long time but its because i rewrite code / take notes. I think maybe one of the best book in ML that i have currently read.
Do you think you can make an updated video of your datascience work flow? Such as project and test directory structure etc.. Your quality has improved a lot, keep up the great work!
Abhishek - your book is one of the best books ever written. Period. To the point. No fluff. LOVE THAT! It should be mandatory for every ML student. I am halfway - and it keeps getting better! Thank you!
Your suggestion is similar to mine. I suggest people to build a usable web demo (or a GIF that is easy to understand if your results cannot be reproduced reliably on other people's webcam or whatever). I think that it all comes down to think like you are the employer or the HR guy. If you can think what other people want from you, you will be able to satisfy them and get a job. Those HR guy and the employer don't know anything about technology. They only care about RESULTS. Can you provide that to them? Can you present it to them in a way they can understand fast and easy? Ask yourself. Show yourself authentically that you can provide results. Fundamentally, it's all about stopping selfishness (thinking only about how your can get a job so YOU can get money). You should think about the VALUE that you have to offer to other people. Think about the goal of the other person then you will be able to manipulate them to your advantage. If you can change your thinking viewpoint fundamentally, you would not need a tip from anyone. You would be the one coming up with the tips!
Very True at end (about Articles), I'll take all your advices, you are really doing a great job thanks Abhishek Sir EDIT: And one more thing I am reading your book that is pretty awesome, it never looks like a sleep during reading.
Haha 😅 I have rarely seen anyone liking a comment! Sometimes I suspect "Abhishek" has deployed a YT bot to like positive comments! (joking) That could save time of the author while also giving the commenters a satisfaction/joy which can drive more traffic! Looks like I found a project which involves some NLP to work on! -> "RUclips Positive Comment Liker" 😂😂😂
Hey! what is overfitting the test data?..i have heard of overfitting in general that shows u fit train data but dont generalize on test data. How is it possible to overfit test data when I am not at all using test data to create model?.
Hey Abhishek, thanks for another amazing and extremely helpful video. Can you do a video on programming? Like if someone wants to improve his programming in Data Science projects, how he/she can go about it. I know you have given it a lot of practice, so I you can tell what all steps someone can follow to gain high proficiency in writing this kind of code, then it will be awesome!
What do you think is the future of Data Science jobs? Do you think there will be a boom for 2-3 years and again an AI winter? Cause many companies are not able to generate revenue via AI
Could you discuss how to write good and understandable code? I have been programming for sometime but still difficult to know if my code is good or not, I think it's very subjective and everyone has his/her own coding style.
Hello Abhishek, Great stuff I brought your book "Approaching any Machine Learning problem" and I completed it. Book is filled with code I really enjoyed it, Thanks !. I have few questions which I am struggling to find answers. Why do we randomize feature selection from a dataset in RandomForest. What happens if we select the features sequentially? If I have 10 million rows, what should be the sample size for my bootstrapped dataset ? What is the difference between attention and lstm with return states ? How do you perform feature selection in Neural Networks? Thanks in advance, Will be waiting for your reply.
Really Nice , I published one blog on ML but I made some mistakes which you mentioned here.I will make changes. Can you please make a video on Image segmentation data like severestal steel Data or Pneumothorax or which you feel better to explain which have multiple options to learn
Can you make a video on explaining how and what to import from projects into jupyter notebook? I have come across various repositories which upload their ipython notebook and do not have a single .py file. Can you also tell the importance of OOP while doing data science projects
Is reading published papers and implementing it and trying to achieve the same accuracy and stats using a particular framework a good idea for ML projects and for the portfolio ?..... Love your informative videos keeps me motivated always to work hard and learn more..
Thanks. Yes! That is a quite good idea! All you need to do after implementation is share it with the world! A very popular example is huggingface's transformers or pytorch-pretrained models repository. :)
How can we implement your information. When ever you start teaching, you start with new libraries, which I never have hear before. You say that- I am not going to teach you about libraries just start using it. When ever I try to learn from you. I find my self stuck with install different libraries.
Have you trued to google individual libraries? Your comment here is not even related to the video. The channel is not for absolute beginners. Try doing the course from Andrew Ng first.
@@amandarash135 When you encounter a new library, before installing it, go to the github page of the library and see the readme. have a vague idea of what the library is doing. try to dive into the code of the library to see the how the components that we use are implemented. when you start doing this, you will understand more about these libs. whenever there is a new library that I use in my video (new to most data scientists), I do talk about it and how it works very briefly. are there some specific libraries that you are struggling with?
@@abhishekkrthakur 😅😅😅 yes! Now I have been trying to understand every pkgs of tensorflow. This is shame full. I don't know yet . Because when ever I did some projects on ml. I have implemented every function on my own. But now I have to deal big libraries, which I don't know fully and what's going on inside these things. Conclusion is I am not able to use libraries until I come to know what is going inside and how does it look like or it is just about Using "tf.randomforest ".
Maybe he is sad after seeing soo many bad ML/AI blogs out there! He wants to improve the DS/ML/AI community by guiding through his YT channel! Thats kind of a cool way! 🙈😆 BTW I love this video! (I meant I really needed it now! 😍)
![BLACK BAG - Official Trailer [HD] - Only in Theaters March 14](http://i.ytimg.com/vi/Du0Xp8WX_7I/mqdefault.jpg)
Great Video, Abhishek Sir...
It was very insightful! Thank you.
Summary:
1. Choose a good project
2. Experiment X 3
3. Write good & understandable code
4. Share your code with good documentation (ReadME)
5. Mention License
6. Package your project
7. Make a good looking web application
8. Make a cool demo
9. Write an article/blog
10. Share it on linkedin/twitter etc
like, subscribe and share? 🤣
@@abhishekkrthakur Good catch !!!!
@@abhishekkrthakur sahi pakde
Can't appreciate much to all the selfless help you are doing for the community. :)
Abhishek : this video is very different from what i usually do
Me: totally , i don't see cover of his book in the background
BTW loved the book
dang, i missed it haha . thank you 💙
i'm currently reading your book. I LOVE IT. I take a long time but its because i rewrite code / take notes. I think maybe one of the best book in ML that i have currently read.
Thank you. Please consider writing a review on Amazon too :)
Abhishek! Thank you for all you do! You are my ML Hero !
Sir, you are down to earth person as helping others
Do you think you can make an updated video of your datascience work flow? Such as project and test directory structure etc.. Your quality has improved a lot, keep up the great work!
I think I can :)
@@abhishekkrthakur would love to see that.
@@abhishekkrthakur I look forward to it too
Woow Abhishek, this is a great eye opener
@Abhishek Thakur : Nice and Useful Information. Thanks
Abhishek - your book is one of the best books ever written. Period. To the point. No fluff. LOVE THAT! It should be mandatory for every ML student. I am halfway - and it keeps getting better! Thank you!
Thank you . Please consider writing a review on Amazon too 💙
@@abhishekkrthakur Done! ;-)
Thanks for the guidance sir :)
Your suggestion is similar to mine. I suggest people to build a usable web demo (or a GIF that is easy to understand if your results cannot be reproduced reliably on other people's webcam or whatever). I think that it all comes down to think like you are the employer or the HR guy. If you can think what other people want from you, you will be able to satisfy them and get a job. Those HR guy and the employer don't know anything about technology. They only care about RESULTS. Can you provide that to them? Can you present it to them in a way they can understand fast and easy? Ask yourself. Show yourself authentically that you can provide results. Fundamentally, it's all about stopping selfishness (thinking only about how your can get a job so YOU can get money). You should think about the VALUE that you have to offer to other people. Think about the goal of the other person then you will be able to manipulate them to your advantage.
If you can change your thinking viewpoint fundamentally, you would not need a tip from anyone. You would be the one coming up with the tips!
Amazing tips sir thank you very much
আপনার এই ভিডিও টা কি যে হেল্প করলো ! অসংখ্য ধন্যবাদ ভাই।
Very True at end (about Articles), I'll take all your advices, you are really doing a great job thanks Abhishek Sir
EDIT: And one more thing I am reading your book that is pretty awesome, it never looks like a sleep during reading.
Thank you very much for sharing.
Will definitely give it a try, Thanks Abhishek sir!
Great content as always!
Thank you :)
Sir really it helped me so much
Thank you so much 😘 😘😘
It would be helpful,if you could share some project ideas! Loved this one!
I am using django and react.js to develop web application for machine learning
thats God mode :D
Thanks for doing this kind of video.
This is priceless for freshers like me 🙏🙏
🤗
This is awesome!
Haha 😅 I have rarely seen anyone liking a comment!
Sometimes I suspect "Abhishek" has deployed a YT bot to like positive comments! (joking)
That could save time of the author while also giving the commenters a satisfaction/joy which can drive more traffic!
Looks like I found a project which involves some NLP to work on!
-> "RUclips Positive Comment Liker" 😂😂😂
would be able to do a CatBoost and LGBM tutorial? thanks and loving your content as always!
Yep. Any preferred datasets?
@@abhishekkrthakur preferably dataset with imbalanced data :) - www.kaggle.com/c/santander-customer-transaction-prediction/
Thank you for your guidance.
Hey! what is overfitting the test data?..i have heard of overfitting in general that shows u fit train data but dont generalize on test data. How is it possible to overfit test data when I am not at all using test data to create model?.
Hey Abhishek, thanks for another amazing and extremely helpful video. Can you do a video on programming? Like if someone wants to improve his programming in Data Science projects, how he/she can go about it. I know you have given it a lot of practice, so I you can tell what all steps someone can follow to gain high proficiency in writing this kind of code, then it will be awesome!
Quite informative and motivating.
Very insightful.
What do you think is the future of Data Science jobs? Do you think there will be a boom for 2-3 years and again an AI winter? Cause many companies are not able to generate revenue via AI
A good video👌
Could you discuss how to write good and understandable code? I have been programming for sometime but still difficult to know if my code is good or not, I think it's very subjective and everyone has his/her own coding style.
Hello Abhishek, Great stuff I brought your book "Approaching any Machine Learning problem" and I completed it. Book is filled with code I really enjoyed it, Thanks !.
I have few questions which I am struggling to find answers.
Why do we randomize feature selection from a dataset in RandomForest. What happens if we select the features sequentially?
If I have 10 million rows, what should be the sample size for my bootstrapped dataset ?
What is the difference between attention and lstm with return states ?
How do you perform feature selection in Neural Networks?
Thanks in advance, Will be waiting for your reply.
Hi Abhishek!! thank you for your videos!
Have you heard about fastai? do you have any opinion about that courses? thanks!
Hi Abhishek, do you use Pyspark for dealing with Big Data for Machine Learning or Deep Learning tasks? or just Pandas?
Good stuff
Really Nice ,
I published one blog on ML but I made some mistakes which you mentioned here.I will make changes.
Can you please make a video on Image segmentation data like severestal steel Data or Pneumothorax or which you feel better to explain which have multiple options to learn
I will make one.
Sir you are inspiration.
Its just your kindness :)
Hi, thanks a lot for such an informative video. Could you please share few example project?
Thank you! So many datasets available. Lemme see if I can come up with something :)
Sure, that would be great.
Can you make a video on explaining how and what to import from projects into jupyter notebook? I have come across various repositories which upload their ipython notebook and do not have a single .py file. Can you also tell the importance of OOP while doing data science projects
Interesting! Please mentor us more! :)
What if we have some unqine project should we post this project on LinkedIn ?
ofcourse! something unique must be shared!
@@abhishekkrthakur if someone copy my project then .
Is reading published papers and implementing it and trying to achieve the same accuracy and stats using a particular framework a good idea for ML projects and for the portfolio ?..... Love your informative videos keeps me motivated always to work hard and learn more..
Thanks. Yes! That is a quite good idea! All you need to do after implementation is share it with the world!
A very popular example is huggingface's transformers or pytorch-pretrained models repository. :)
How can we implement your information. When ever you start teaching, you start with new libraries, which I never have hear before. You say that- I am not going to teach you about libraries just start using it. When ever I try to learn from you. I find my self stuck with install different libraries.
Have you trued to google individual libraries? Your comment here is not even related to the video. The channel is not for absolute beginners. Try doing the course from Andrew Ng first.
@@abhishekkrthakur yes I have done. He teaches us about algorithm and how to use it. But not about much libraries. I think you better know
@@amandarash135 When you encounter a new library, before installing it, go to the github page of the library and see the readme. have a vague idea of what the library is doing. try to dive into the code of the library to see the how the components that we use are implemented. when you start doing this, you will understand more about these libs. whenever there is a new library that I use in my video (new to most data scientists), I do talk about it and how it works very briefly. are there some specific libraries that you are struggling with?
@@abhishekkrthakur 😅😅😅 yes!
Now I have been trying to understand every pkgs of tensorflow. This is shame full. I don't know yet . Because when ever I did some projects on ml. I have implemented every function on my own. But now I have to deal big libraries, which I don't know fully and what's going on inside these things. Conclusion is I am not able to use libraries until I come to know what is going inside and how does it look like or it is just about
Using "tf.randomforest ".
@@amandarash135 you don't need to understand everything ;)
Really waiting for that RL video(s) ..xD
me too :D :D
hopefully soon. too many things in list :)
SUPER INTERESTING
Thank you!
@@abhishekkrthakur Sir, you are a true inspiration of my Data Science Career. Thank you so much 💓
Buddhadeb / Aaroha
I bought the book.....will I be considered access to your Github repo almostnlp :)
Announcement soon :)
😀😀😀
first view
yay! :)
Your name reminds me of Tupac shakur
haha :D
You look Sad
why 😂 im not 😎
Maybe he is sad after seeing soo many bad ML/AI blogs out there!
He wants to improve the DS/ML/AI community by guiding through his YT channel!
Thats kind of a cool way! 🙈😆
BTW I love this video! (I meant I really needed it now! 😍)
thank you :)