THE TREE OF DECISIONS

What if machine learning could begin with a question you already ask every day?
Should this request be handled now or later? Does this case need human review? Which detail changes the next step?
The Tree of Decisions turns those familiar moments into a gentle, visual introduction to one of the most understandable ideas in artificial intelligence: the decision tree.
Written for nontechnical readers, this book begins before the formulas. It starts with ordinary choices, clear conversations, and small hand-built examples. Step by step, you will see how a difficult problem becomes a sequence of questions; how examples become data; how roots, branches, and leaves create a visible path; and how a model learns to choose useful splits.
When the mathematics arrives, it has a purpose. Counts, proportions, Gini impurity, entropy, weighted averages, information gain, error, depth, validation, overfitting, and pruning are explained from first principles in plain language. No programming experience is required.
But a readable model is not automatically a trustworthy model. The book also examines biased labels, proxy variables, data leakage, unstable thresholds, distribution shift, uncertain cases, and the need for human oversight. Applications in customer support, education, healthcare support, finance, agriculture, manufacturing, and public services show both the value and the limits of tree-based reasoning.
Inside, readers will:

• build a complete decision tree by hand;
• follow new cases from root to leaf;
• compare candidate splits;
• understand why deeper is not always better;
• learn when to use, adapt, or avoid a decision tree; and
• practise with recall maps, chunking exercises, a capstone project, a glossary, and a thirty-day plan.