Book Image

Solutions Architect's Handbook

By : Saurabh Shrivastava, Neelanjali Srivastav
Book Image

Solutions Architect's Handbook

By: Saurabh Shrivastava, Neelanjali Srivastav

Overview of this book

Becoming a solutions architect gives you the flexibility to work with cutting-edge technologies and define product strategies. This handbook takes you through the essential concepts, design principles and patterns, architectural considerations, and all the latest technology that you need to know to become a successful solutions architect. This book starts with a quick introduction to the fundamentals of solution architecture design principles and attributes that will assist you in understanding how solution architecture benefits software projects across enterprises. You'll learn what a cloud migration and application modernization framework looks like, and will use microservices, event-driven, cache-based, and serverless patterns to design robust architectures. You'll then explore the main pillars of architecture design, including performance, scalability, cost optimization, security, operational excellence, and DevOps. Additionally, you'll also learn advanced concepts relating to big data, machine learning, and the Internet of Things (IoT). Finally, you'll get to grips with the documentation of architecture design and the soft skills that are necessary to become a better solutions architect. By the end of this book, you'll have learned techniques to create an efficient architecture design that meets your business requirements.
Table of Contents (18 chapters)

Evaluating ML models – overfitting versus underfitting

In overfitting, your model fails to generalize. You will determine an overfitting model when it performs well on the training set but poorly on the test set. This typically indicates that the model is too flexible for the amount of training data, and this flexibility allows it to memorize the data, including noise. Overfitting corresponds to high variance, where small changes in the training data result in big changes to the results.

In underfitting, your model fails to capture essential patterns in the training dataset. Typically, underfitting indicates the model is too simple or has too few explanatory variables. An underfitted model is not flexible enough to model real patterns and corresponds to high bias, which indicates the results show a systematic lack of fit in a certain region.

The following graph illustrates the clear difference between overfitting and underfitting as they correspond to a model with good fit: