Customer reviews

Rating distribution

4.6

(5 Ratings)

5 star 60%

4 star 40%

3 star 0%

2 star 0%

1 star 0%

Steven Fernandes Apr 12, 2024

This insightful book dives deep into the essentials of machine learning and artificial intelligence, stressing the importance of data quality, expert involvement, and ethical considerations above complex algorithms. It offers practical advice on data handling, including cleaning, labeling, and augmentation, and provides a thorough guide to synthetic data generation with Python. The book is particularly strong on addressing bias, presenting effective strategies for creating fair and equitable AI systems. It rounds off with a critical look at the ethics and responsibilities in AI development. A must-read for anyone in the field, blending technical guidance with a strong ethical framework.

Amazon Verified review

Advitya Gemawat Apr 18, 2024

Going through the book honestly made me walk back memory lane when I took a 'Practice and Applications of Data Science' class back in sophomore year of college.- 📈 **Data Preprocessing**: The book covers various Data Imputation techniques used to handling missing data. It also discusses different ways to detect and handle outliers, including Z-score, IQR method, and the corresponding use of scatter plots and box plots.- 🧠 **Model Selection**: There's a nice walkthrough of various traditional #ML models like Linear & Logistic regression, Decision Trees, and SVMs, along with their assumptions, pros & cons, and code snippets for using these models with scikit-learn. The bullet points in Chapter 9 can especially be good to use for interview prep for college students- 📊 **Feature Engineering**: The book covers foundational techniques like binning, log transform, one-hot encoding, and interaction features, along with overarching strategies such as Active Learning, Weak Supervision, and Semi-supervised learning.But my favorite part of the book was the **Data-Centricity** aspect. Quite frankly, I initially found the book title a bit vague as data-centricity in ML sounded obvious to me. But this is one of the few books I even ended up reading the 'Foreword' of, which I've usually skipped. And here's why -The Foreword combined with the last chapter had an well-intentional storyline of introducing the philosophical foundations of how data was collected throughout history, *how* (not just why) data is the main and most important fabric of the ML lifecycle, and even touching upon Responsible #AI at the end.This storyline can be quite beneficial for folks newly getting into the ML space - for a traditional model or an LLM, groundedness in high-quality data is the most important and effective driver of the lifecycle.The book reads like a more relevant resource for early-in-career professionals looking to enter the AI industry, folks looking to strengthen their fundamentals and use for interview prep, and college students with a basic programming background.

Amazon Verified review

H2N Apr 15, 2024

The book discusses about improving data quality in order to get robust, fair, and interpretable ML models. It challenges the traditional focus on algorithms by shifting towards data-centric methods. This book provides lots of practical examples using modern techniques, emphasizing collaboration between data scientists and domain experts to refine data quality, making it an essential guide for advancing in AI.

Amazon Verified review

Om S Apr 12, 2024

When I first started out in machine learning, I was completely absorbed in tweaking algorithms and tuning models, always hunting for that slight improvement in accuracy. However, I quickly realized that no matter how much I optimized the models, the real barrier to superior performance was the quality of the underlying data. This revelation shifted my focus entirely and led me to explore data-centric approaches in machine learning—a journey that transformed my projects.This book provides a thorough guide to data-centric machine learning, emphasizing the critical role of high-quality data over merely adjusting models. It introduces the impactful concept of 'small data' and how it can revolutionize ML/AI projects. Through practical strategies for enhancing data collection, labeling, and augmentation, the book ensures readers can refine their datasets effectively. It explores the importance of synthetic data and the necessity of expert involvement in data labeling, equipping professionals with tools to boost their ML practices. The book also addresses how to identify and eliminate biases and stresses the ethical dimensions of machine learning, offering a comprehensive view of both the challenges and solutions in the field. Ideal for data science professionals and enthusiasts, this book serves as an essential resource for anyone eager to prioritize data quality in their machine learning projects.

Amazon Verified review

Amazon Customer Mar 24, 2024

The book adeptly explores the critical importance of data quality in ML projects, providing practical insights and techniques for creating reliable datasets. From data labeling to synthetic data generation, the book equips readers with the essential skills to enhance their ML practices. With a focus on ethics and responsibility, this book serves as a valuable resource for data professionals and ML enthusiasts seeking to elevate their projects through data-centric approaches.

Amazon Verified review

Data-Centric Machine Learning with Python: The ultimate guide to engineering and deploying high-quality models based on good data

What do you get with eBook?