What is Machine Learning?

Introduction

Machine learning builds programs whose behavior is shaped by data. You supply examples; an algorithm adjusts internal parameters so predictions on new inputs are useful.

Understanding the topic

Rule-based code vs learned models:

• Traditional: engineers write explicit if/else and formulas.
• ML: engineers choose features, loss, and algorithm; parameters are learned.
• Generalization to unseen rows is the goal — memorizing training rows is failure.
• Data quality, coverage, and feature design usually beat algorithm trivia.

Informative example

Baseline classifier in scikit-learn:

from sklearn.datasets import load_wine
from sklearn.model_selection import train_test_split
from sklearn.ensemble import RandomForestClassifier
from sklearn.metrics import accuracy_score

X, y = load_wine(return_X_y=True)
X_tr, X_te, y_tr, y_te = train_test_split(X, y, test_size=0.25, random_state=7)

clf = RandomForestClassifier(random_state=7)
clf.fit(X_tr, y_tr)
print(round(accuracy_score(y_te, clf.predict(X_te)), 3))

1.0

Execution workflow

Best practices

• Always establish a dumb baseline (majority class or mean predictor).
• Version datasets and preprocessing code alongside model weights.

Summary

ML automates pattern discovery from data while you own problem framing, metrics, and deployment.