Supervised Machine Learning Example in Python - KAGGLE COMPETITION - Code is Provided

Подписаться 26 тыс.

Просмотров 7 тыс.

50% 1

Solve your first Kaggle Competition! From a csv file all the way to making predictions and deploying your results. Full end-to-end Tutorial on Machine Learning in Python. We start by explaining the Machine Learning Process. Then, we move on to the Data pre-processing phase where we clean and transform our data. We show some methods on how to identify the most important variables.
After that, we run the model and make predictions. Every time we run a model, we submit it into Kaggle to get our score on the evaluation dataset. Then, we go over a few methods on how to improve our results & predictions. We provide the raw data and the code! Hope you enjoy!
Data Analytics Course Link:
ipidata.teachab...
Raw Data and Code:
github.com/Pit...
Part 2:
• Machine Learning - Dea...
Other Supporting Videos:
Video 1 - Down and Install Python - Numpy Tutorial:
• How to learn Python? -...
Video 2 - Pandas Tutorial:
• Complete PYTHON Tutori...
Video 3 - JOINs and UNIONs Tutorial:
• How to Merge DataFrame...
Video 4 - Data Visualizations with MatPlotLib:
• How to create Data Vis...
Video 5 - Data Visualizations with Seaborn:
• Complete Seaborn Tutor...
Video 6 - Machine Learning Example - Regression:
• Machine Learning Tutor...
Table of content:
What is machine Learning?
How to run machine learning in python?
Supervised machine learning example in python
What is the machine learning process
How to clean data in python?
How to do data pre-processing python machine learning
How to deal with outliers in python?
How to investigate the distributions in python?
How to do feature engineering in python?
How to find the most important variables in python?
What is a machine Learning regression model and how it works?
How to run machine learning classification model in python?
How to optimise a machine learning model in python?
How to Run XGBoost in Python?
Yiannis Pitsillides on Social Media:
/ pitsillides91
www.instagram....
www.pinterest....
www.facebook.c...

Опубликовано:

20 окт 2024

Ссылка:

Скачать:

Готовим ссылку...

Добавить в:

Мой плейлист

Посмотреть позже

Комментарии : 24

@YiannisPi 2 года назад

What do you think about the ML videos? Shall I keep doing them?

@sagarshelar3927 2 года назад

Yesss🤝🔥🔥

@mdsamiulislam2522 2 года назад

U should do

@marcoschwarz4640 2 года назад

100%

@adilmajeed8439 2 года назад

Please keep continue these videos, this is quite unique as you are bringing the machine learning stuff along with the visualization tools like Power BI

@rolandnwatarali9716 2 года назад

Your Machine Learning Videos is not only Value adding but gives a representation of how real world data challenges are being handled. Please continue making these type of videos. Can i suggest you bring in a bit of Real world datasets (Twitter for example) in to the Picture. And work around different data stages (From Cleaning the Dataset - Performing EDA's - Building Machine Learning Models - Final Deployment in PowerBi). Many Thanks for the relentless effort being Put to create the Value adding Videos. Gracias !!!!

@udaynj 9 месяцев назад

Great tutorial. You had just the right speed while speaking and explained everything really clearly. Thanks again! Really great job!

@canerkoldemir3853 Год назад

That was so clear and understandable thanks to your knowledge

@syedhaideralizaidi1828 2 года назад

You are literally the best out there, teaching line by line putting in so much of your time to create detailed documented notebooks. Such a treat to learn from you. Kudos to you. ❤️ When will the second part be uploaded?

@YiannisPi 2 года назад

Thanks Syed, Appreciate this!

@c-historia Год назад

this is great! 🎥

@KenyanPigment 2 года назад

Big love from Nairobi Kenya❤️

@idrees2516 2 года назад

ofcourse please continue it

@chimaaustine2200 Год назад

Thanks a lot for the awesome resource. You have further simplified the process for me. However, I do have a question/scenario regarding how historical datasets are treated in ML. If I have a dataset of customers and their purchases in a shop over a period of time. 1) Do I necessarily need to flatten the dataset in someway during processing stage ?. If yes, How would you recommend? 2) If I use the dataset as is without any form of flattening, this means "customer_id" row won't be unique obviously, would this cause problems in the accuracy of the model ? I hope the question is clear though. Thanks

@iQualizer 2 года назад

Thank you very much for your work. Could you please share the link to the second video?

@YiannisPi 2 года назад

Will be posting it early next week! Apologies for the delay

@iQualizer 2 года назад

@@YiannisPi Thank you!

@qaboahene 2 года назад

Nice one. Maybe next time you could explore using other methods to handle the missing values? I want to see if it’ll help with the model in general.

@Manapoker1 Год назад

Really nice video, I learned a lot, than you!

@Satramsarabdial 2 года назад

I think there is an error in the code. When we converted the string data into (str) we converted nans to strings, so when we fillna with mode they are never replace because they are now strings not nans. raw_data_clean.isna().sum() will show no nans but there are. If you run raw_data_clean[raw_data_clean['Deck'] == 'nan'] you will get 200 nans. This is the case for all the string data. Yannis might have addressed and maybe I missed it>

@randb9378 2 года назад

Great video! i have a question related to the exercise above. When we have a binary value (e.g True False) and a continuous variable (Income), can we convert the binary value to zero and one and then calculate the correlation or is that bad practice? Thanks once again - continue the great work