In these unique times of economic disruption bring unique opportunities for learning new skills and preparing for the upswing of new employment after all is said and done.
Machine learning (ML) will be particularly useful given that virtually all businesses are converting to digital workflows and service-based products.
The smooth flow of data operations will be vital to succeed in this new world, and ML is the best way to bring the power of automation to the infrastructure and processes that keep users engaged and satisfied with their digital services and connected devices.
While much has been written as to how machine learning and the broader field of artificial intelligence (AI) will make many of today’s jobs obsolete, this view overlooks the more salient fact:
Any jobs that are lost will not be lost to ML, but to people who know how to leverage machine learning to become more productive.
With ML, a single IT professional can automate all of their rote, repetitive and least productive daily tasks so they can concentrate on higher-order strategic issues.
This in turn will produce measurable benefits to the organization’s bottom line through greater performance, lower costs, opening up of new markets, development of new products and many other ways.
These gains can be easily measured, tracked and sourced to the employee who is responsible — the person whose daily tasks are being done by machines.
The challenge, of course, is to acquire the skillsets needed to leverage ML effectively.
Can this truly be learned with relative ease and at low cost? And can it be done at home, in one’s spare time?
The answer to both questions is a qualified “yes.”
All that is needed are the right resources that explain this technology simply and effectively.
The following eBooks are available, affordable, and provide a solid foundation for beginning a new career, or enhancing an existing one, with machine learning.
Note: Want something more advanced? "Online Learning: Top 5 eBooks for Machine Learning Experts.)
Python is one of the most popular ML languages, so it is a good place to start. This eBook provides the basics to writing clean, concise code in Python 3.
It stresses object-oriented programming (OOP) that can be applied to entry-level data science, visualization, web apps and other projects. It also provides guidance on how Python can be used to automate data-to-day tasks using predictive machine learning.
Instruction is presented as a workshop, with step-by-step learning tools that stress the practical side of Python programming, not heavy, abstract theorizing. It also allows students to learn at their own pace, running through a single exercise per day or cramming an entire course over a weekend.
All the while, new skills are being added as the student develops real-world code, and there is even a certificate upon completion that can be shared and verified.
To work with ML, it is important to understand how it works, and more importantly, how it differs from traditional software.
This book takes the basics of Python and applies them to ML in relation to real-world concepts and problems. Along the way, you’ll learn how to apply regression and classification techniques, as well as predictive analysis using decision trees and random forests.
As you progress, you’ll go from simple tasks to building fully intelligent applications featuring advanced tools, like k-means and mean shift algorithms, and progressing all the way up to deep learning and artificial neural networks (ANN).
Upon completion, you should be able to launch your own ML applications into real-world scenarios.
pandas is the most popular library in the Python universe, employed by data analysts the world over to manipulate and organize digital information.
This book provides an in-depth look at pandas’ numerous tools, such as multi-indexing, data structure modification and sampling. It also enables both intermediate and experienced practitioners to apply data insights to multiple domains, including Bayesian statistics, predictive analytics and time-series analysis.
Upon completion, readers will understand how to use pandas on complex data sets for more efficient analysis and more accurate results.
It also includes instruction on how to prepare interactive business reports using the popular Jupyter notebook.
Now that the basics of programming and data manipulation are understood, it’s time to get to the heart of machine learning.
This book offers the means to master the frameworks, models and techniques that allow machines to actually "learn" from data. It features instruction on how to use scikit-learn and TensorFlow 2.0 for both ML and deep learning, as well as ways to apply these concepts to image classification, intelligent applications and other projects.
As well, readers will learn the latest methods for building and training artificial neural networks, GANs and other models, as well as the best ways to evaluate their operations and optimize performance.
Additional lessons cover the ways to use regression analysis to predict continuous target outcomes, as well as using sentiment analysis, a newly developed sub-field of Natural Language Processing (NLP), to evaluate text and social media.
The book covers both the principals of Python-based machine learning as well as the practical applications using clear explanation, visualization and working examples, allowing students to develop real-world models and applications by themselves.
For those who have now mastered the basics of ML, this book provides the fine-tuning necessary to create enriched models, streamlined processes and more elegant code.
With how-to guides covering key Python tools like pandas, scikit-learn, Featuretools and Feature-engine, readers will discover the ins and outs of working with continuous and discrete datasets, as well as transforming features from unstructured data.
Key “recipes” also illustrate how to automate feature engineering to simplify complex processes using a range of techniques like box-cox transform, power transform and log transform, and then apply these processes to machine learning, reinforcement learning and NLP.
As well, readers will learn key techniques like how to impute missing values, how to encode categorical variables, how to quickly and easily extract insights from text, and how to develop new features from transactional and time series data.
With any complex learning endeavor, the best teacher is experience. Nevertheless, we all have to start somewhere.
For those facing serious downtime at home during the present crisis, there are two ways to endure: stare at the walls while the hours tick by, or take the initiative to be in the best position to excel in the post-virus economy.