Approaching (Almost) Any Machine Learning Problem

by Abhishek Thakur

the book is 3+ years old now and there is currently no plan to update it. Its better to just get free PDF from github This is not a traditional book.The book has a lot of code. If you don't like the code first approach do not buy this book. Making code available on Github is not an option.This book is for people who have some theoretical knowledge of machine learning and deep learning and want to dive into applied machine learning. The book doesn't explain the algorithms but is more oriented towards how and what should you use to solve machine learning and deep learning problems. The book is not for you if you are looking for pure basics. The book is for you if you are looking for guidance on approaching machine learning problems. The book is best enjoyed with a cup of coffee and a laptop/workstation where you can code along.Table of Setting up your working environment- Supervised vs unsupervised learning- Cross-validation- Evaluation metrics- Arranging machine learning projects- Approaching categorical variables- Feature engineering- Feature selection- Hyperparameter optimization- Approaching image classification & segmentation- Approaching text classification/regression- Approaching ensembling and stacking- Approaching reproducible code & model servingThere are no sub-headings. Important terms are written in bold.I will be answering all your queries related to the book and will be making YouTube tutorials to cover what has not been discussed in the book. To ask questions/doubts, please create an issue on github Subscribe to my youtube

Genres: Computer Science, Artificial Intelligence, Technology, Technical, Academia, Nonfiction

300 pages, Kindle Edition

Published July 4, 2020

Reviews

[Guido Fawkes · Rating 2 out of 5]

There is so much hype around this book. All over the social media, you can find people posting photos of the book as just unboxed from mail and claiming the book will render them some kind of demi-god of machine learning.

There are a few critic voices, too, but they are put into silence by the very aggressive behavior of the author. I read the book from cover to cover and I can tell you that it is fool's gold backed by a cargo cult.

The book is simply normal content, nothing so special, the same stuff and code you can freely find on blogs and github all over the internet provided by many amateurs and enthusiastic people working in data science. The author is clearly working everyday with data science stuff and he is providing some of his knowledge, though not in a very organized way: the contents are a kind of wild spaghetti data science, a stream of consciousness in an analyst's mind. Don't be amazed if you cannot understand what the author wants to tell you in a chapter because within each chapter, which acts as a content box, the contents are very dispersive and not organized at all. That's all folks, nothing special in it.

You can also find a lot of code that the author claims you have to digit by yourself so you can learn. You have to find the datasets by yourself. That's so much cargo cult. In reality the code is not well explained because the author doesn't explain it but for some comments in the code itself.

Buy at your own risk, if you need a book on doing data science you can find better content buying Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow by Aurélien Géron.

[Amirali · Rating 3 out of 5]

Good if you are getting into machine learning. The book does not really go into newer topics like deep learning, attention models and reinforcement learning.

[Jack Leitch · Rating 4 out of 5]

Very pragmatic

[Ishaku Ibrahim]

I am new to machine learning. I have struggled a bit to grasp concepts but this book instantly change everything for me. From cross validation to evaluation metrics and so on ....just wow. The simplified approach is a big plus for any reader. Thank you Abishek Thakur

[Rick Sam · Rating 3 out of 5]

A Hands On Approach to almost (all) Machine Learning Problem

Last chapter, included Docker - wonderful.

Suggested first reading for everyone.

[Aadil Srivastava · Rating 5 out of 5]

A great read!!

If you are a Machine Learning practitioner and you already are familiar with the basics as well as the theoretical part of ML/DL, then this book is gonna is very good read for you. It's not a typical ML book that will start with thr basics like what ks ML, Linear Regression, etc. instead it will you to gain maximum out them.

It has already helped in recognising gaps in my understanding of practical ML.
Some of the small-small ideas are so brilliant that it gave me a new perspective and I'm wondering it would have been so much fun I have used these things when I was solving a similar bussiness problem.

Thanks a lot Abhishek for putting so much of your effort.

[Arpit Vijay]

its a great book

[Nishant Bhadauria · Rating 5 out of 5]

A really remarkable effort

This is an hands on book and not a theory book as already mentioned by author.The points that make this book different

1. Using comprehensive code with good explanation and all necessary comments
2. Not using standard dataset like iris cars etc and utilising bigger Datasets from kaggle
3. Touching almost everything that you encounter while building a model. Also adding on touching distributing your model using flask and docker
4. Covers NLP too including transformers which many of starting ML books choose to ignore.

For me it is a good reference guide if you brush up again and again.

[Yassine Alouini · Rating 4 out of 5]

Worth the money and time!
Only the computer vision chapter felt a little bit rushed. Also, I wish there was a little bit more theory but I guess you can't have everything but almost everything. 😁

[Varun Gadre · Rating 5 out of 5]

Great book to get started with ML.

Its a nice bridge between the knowledge part and the applied part of ML. The codes are simple to understand with the help of comments. The book also helped in understanding your Youtube tutorials and vice-versa. The book has really got me started on approaching problems on Kaggle. Looking forward to future books in the series.

[Rakesh · Rating 5 out of 5]

Best book ever read on ML. Very practical.
Would recommend to anyone who are new or experienced in the field. Detailed and to the point of explaining things. One could surely get a job in the field of Data Science if he understands and implement everything given in the book.

[Misal · Rating 5 out of 5]

If you are a practitioner of ML, I'd suggest you buy this book with your eyes closed. The author doesn't go extremely deep into the theory behind most of the algorithms. As the name states, this book is very practical. It is focused on DOING more than LEARNING.
As far as the book the page quality is very nice too, and they have a glossy finish.

[Vladimir Slaykovsky · Rating 4 out of 5]

Solid beginner's practical introduction to ML. Some weird swings of authors thoughts and lots of duplicated code.

[Suraj Wate · Rating 5 out of 5]

Amazing book if you like coding.

Best thing I liked about the book is that it took examples from Kaggle to understand concepts. Covered all important concepts.

[Budi Susanto · Rating 4 out of 5]

Even though program code can now be generated from GenAI, this book still provides important basic insights for anyone who wants to learn more about machine learning.