Approaching (Almost) Any Machine Learning Problem [Print Replica] Kindle Edition

Well, I kind of expected this brochure to be an overview but I didn't expect it to be that shallow and chaotic.The author is jumping from statement to statement, from t-SNE to cross-validation, Sturge's rule and God knows what else all these by page 27.There are no instructions on how to obtain the datasets (I know how and where but that's not the point, we strive for reproducibility here and we are not expected to 'believe' in certain things like in Santa or Yeti).The phrases 'as you can see on this graph' are nothing but funny. Where did that come from?Perhaps, it is well worth 9$ but no more than that. I will stick to the book by Aurelien G.To sum it all up, I will vote for kernels and medium posts but this...it ain't a book :)I can't teach things but I am not trying to. Perhaps we should stick to stuff we are good at?

Kaggle has a strange reputation within the data science community. On one hand it's a great source of innovation in a range of sub-fields and when solving a similar problem to an existing Kaggle competition seeing how it was approached by high ranking teams is very valuable. On the other it is a distorted version of what data science actually is in the real world. Usually the (clean) data is provided to you in Kaggle whereas sourcing, collecting and cleaning data is normally a big chunk of a working data scientists life. Finally the approaches in Kaggle competitions are often all about squeezing that tiny improvement out of large numbers of ensembled models. In the real world concessions towards speed, simplicity and interpretability have to be made.The author Abhishek Thakur was the first to achieve GM level across all 4 categories on Kaggle (competitions, kernels, datasets and discussion) . Even a single GM level is an exceptionally difficult task requiring immense amounts of time and skill. My worry going into this book was who it was aimed at and what its purpose is; is it just about doing well on Kaggle or will people who work in industry learn something valuable? Is it aimed at advanced modellers who are looking to become truly elite or would someone with a more general background gain useful knowledge?I am pleased to state that this is a book which is very valuable for the working data scientist and the keen Kaggler. The real value is how it allows us to see how a highly skilled predictive modeller approaches new problems. The book is made up of 13 chapters;ch 1 - Setting up your working environmentch 2 - Supervised vs Unsupervised Learningch 3 - Cross-validationch 4 - Evaluation Metricsch 5 - Arranging Machine Learning Projectsch 6 - Approaching categorical variablesch 7 - Feature engineeringch 8 - Feature selectionch 9 - Hyperparameter optimisationch 10 - Approaching image classification & segmentationch 11 - Approaching text classification/regressionch 12 - Approaching ensembling and stackingch 13 - Approaching reproducible code & model servingWhile Kaggle is great at discussing a highly placed final entry, the value of this book is a walk through of the steps taken towards a solution; ch 1 on Setting up your working environment, ch 5 Arranging a machine learning project and ch 13 on Reproducible code and model serving I found particularly valuable learning some neat tricks on laying out a project. These are all valuable topics which can get lost when we ask "How did the solution work" as in reality the final answer involved lots of iteration lost in just seeing the final product. I really liked the way the author's projects were laid out using a config file and model_dispatcher file allowing for quick modification of which algorithm to use. I had not come across this before but it's a great idea which speeds up the iteration of the modelling process. The other elements that I think makes this book such a great learning solution for people beginning their data science journey is that it shows mistakes which the author then discusses in depth. Finally we see many examples of simpler models beating more complex ones a lesson that is hard to accept when you are starting out and keen to apply XGBoost or Neural Networks to all the things.This book is not just for beginners however - even as someone who has worked as data scientist in industry for a number of years I learnt a great deal from the chapters on dealing with categorical variables, feature engineering and feature selection. As the author notes there are other sources for these solutions but they are spread out across numerous blog posts and forums - having them in a book makes things much easier. I work in customer analytics so when building predictive models a lot of time in spent on feature engineering and feature selection - I learnt a couple of tricks which will be valuable for new projects at work. The book even includes a section on using embeddings on tabular data - a neat approach not widely used in my experience.Finally the book amazingly includes chapters on computer vision problems using PyTorch for classification and image segmentation and nlp using a range of approaches of increasing complexity from bag of words through word2vec to and LSTM and finally a BERT model. The author rightly skips over the complexities of how a CNN or LSTM and Transformer work, but gives enough of a description to get a sense of what is going on. Again the author emphasises the valuable lesson of starting with simpler models and approaches and only then increasing the complexity with constant comparison to a baseline. The author hints (perhaps jokingly) he is considering work on similar books on Computer Vision and NLP - I hope the success of this book encourages him to seriously consider doing this.It is an amazing achievement that the author has created a book which allows the reader to build strong models in a such broad range of domains. The book is well written with the code in particular being excellent. There were one or two spots where the written phrasing was a little hard to follow but these were rare and overall I enjoyed the writing style. The book is eminently practical so the reader will need to find other sources for the theoretical workings of the algorithms used as they gain more experience. Given the breadth the book achieves this is perfectly acceptable. Finally a small technical issue I had with the Kindle version was the lack of a table of contents accessible via the Kindle menu. Not a big thing but does make navigating the book a little trickier than it needs to be.Overall this is an excellent book full of hard won wisdom from a very talented data scientist and educator. I would happily have paid 4 or 5 times its current price and still been very happy with my purchase. I will be highly recommending this book to friends and colleagues who work (or hope to work) in the field.

I believe that there's a fairly large demographic of people in the community who have probably taken some machine learning course in college. Learning how to program and implement algorithms from scratch is important since this gives us an idea of the inner workings of the theory and mechanics behind the libraries that are implemented in practice. However, I am guessing that it is likely that most of the datasets previously used are quite plain and probably not that difficult or interesting (e.g., the iris dataset). In other words, our experience in academia leaves it quite unprepared regarding how to take the next step.I think that this book does a great job of guiding those who are likely trapped in the academic experience of machine learning, where our work is centered around class assignments or projects that are too cookie-cutter and poorly organized. This book provides some guidance by answering many left-over questions regarding how we can understand a system/methodology that will allow us students to go from homework assignments and towards approaching real-world problems in a professional manner.

This book, written by a legendary Kaggle competitor, is filled with great practical knowledge and code examples (in Python) for approaching machine learning problems. The material is explained simply and clearly. Upon reading through the book I quickly began filling in gaps in my knowledge (fundamental techniques or concepts that I did not know as well as I should). I also learned a lot by looking at the Python code examples, which show many good ways of doing things in Python that I had not known about. Another good aspect of the book is the choice of datasets that are used when illustrating various techniques.

I like this book for its brevity! It is an excellent book for folks who already are aware of various techniques theoretically but kind of hard to pin point where and how to implement them. This book definitely serves the purpose for getting one thinking about how to do it. Although, it is not a detailed outlook (and neither do I expect as there are so many more books covering that) on each topic it definitely is a great reference when you’re stuck on a problem and thinking where to ponder next. The code snippets are excellent as they explain how to make a certain change to a dataset, problem etc and has few tricks as well which I really appreciate. Of course, it’s not comprehensive but again the title itself say approaching any problem and the user is supposed to know a thing or two about ml before taking this book. I would recommend this book for all those trying to quickly spin up machine learning problem and build on top of the results.

It is just a collection of notes and codes that we keep in some random text file, just to resuse it in future. It doesn't give you in depth knowledge of whats and hows of things written in it. Or just maybe I expected something else from this and it turned out to be a simple book of codes.Anyways, it would have been easier just to create a github repo with proper readme files and all, instead of writing a whole book for it.

I am a Data Scientist/ AI Engineer working for the last 3 years in the industry. This is my honest review -As the book name suggests this is the best book that you could get on practical machine learning and data science. The author has written the book with great dedication and the explanations at each section is amazing.I really liked the flow of the book, from basic concepts to covering images and text problems, advanced concepts like Entity embeddings and what not!The codes are written perfectly and in such a way that it develops a great habit of writing good quality code for any data scientist or enthusiast. The author has made sure that the code is reusable and can be taken into production.P.S. - Don’t think too much, just go ahead and buy it!!! It’s Amazing!

I would describe myself as an experienced machine learning practitioner and found this book the most useful I have read in years. I really liked the opportunity to understand why and how the author had approached the particular issues that occur in almost all real world machine learning / data science projects and his detailed approach to items such as crossfold validation.The book consists of text explaining why the author has approached a particular issue in the way that he did together with copious, easy to read and well annotated code. The two combined provide a fascinating insight into how one individual tackles these kinds of projects. There some gaps, such as how do you measure model drift in a production environment, but despite many years experience, I learnt something in almost every topic that the author covered.I'd say the reader needs 2 or more years machine learning / data science experience, so that the reader is familiar with the issues that the author is tackling and the reason that they are chosen. They should also be happy reading python code so they can get the maximum benefit of the author's approach.This is not an academic textbook and contains no theory, but for someone trying to improve the way they tackle real world problems I'd highly recommend it.

It's been 11 days since I got this book. I have refreshed so many topics (with code) and learned many new. I wish to lay out a few points which helped me, and I can carry from the book, but before that, thank you Abhishek Thakur for placing things in order. And trust me, I am waiting for the book you mentioned on page-271.-> Page-3 If you didn't code, you didn't learn. I think this is the best way to learn. I went through a theoretical module on machine learning, and initially, I wasn't getting much, and the moment I started coding for an assignment of the same module, every bit of the module started making sense that's so common, right?-> This book has one thing particular that is code with concise explanations. I love the way the author has derived us throughout different chapters. It helped me to refresh certain topics in order and adapt to the new ones quickly.-> Since I was familiar with basics, I jumped to page 185 and finished till 271, I know that's the wrong way, but I don't know why I did that :-p later, this book forced me to go through the starting chapters as well, since it is well written and organized.-> I would suggest anyone with basic knowledge of machine learning concepts must go through this book and start addressing practical problems.

I don’t normally buy books related to coding for the main reason that they usually become outdated quickly and end up taking up space on my bookshelf, never to be opened again.This book is different though. This is not about learning a specific package or for doing a specific task. This book aims to cover the fundamentals of machine learning regardless of what the task is. The (Python) code examples are easy to follow and use packages with mature APIs (e.g. scikit-learn, pandas, PyTorch) so the code examples will still work for years into the future. It also serves as a great reference for things which are often succinctly explained in a couple of paragraphs (e.g. the explanation of the AUC metric was probably one of the neatest I’ve seen). The other great thing about the writing style is that the theory and equations are intentionally kept to the bare minimum (you can find all of this elsewhere) so that you can get to building things as quickly as possible.If you are new to ML you should buy this book if you:> Have completed some MOOCs and are getting started with your own projects or Kaggle> Are wondering how that ML code that you copied on Stack Overflow/GitHub actually works> Prefer to learn quickly by doing rather than sitting through weeks of classesIf you have ML experience you should buy this book if you:> Started your ML journey through deep learning (e.g. fast.ai) and you want to firm up your ML fundamentals> Are coaching/mentoring less experienced colleagues and you need some succinct definitions/code examples of concepts like cross-validation or a certain metric> Want to improve your “coding style”, e.g. write clearer more maintainable codeI’d definitely recommend this book and it’s something I’ll probably find myself flicking back through for a while. I also really like the cover which has an awesome design and comes in a matt finish which feels nice in the hands while reading. Originally I thought it would pick up fingerprints very quickly but after owning it for several weeks it still looks great.