Foundations of AI and Machine Learning

In traditional programming, humans provide rules and data to get an answer. In Artificial Intelligence (AI), we aim to mimic human cognitive functions. However, AI is a broad umbrella. Machine Learning (ML) is a specific subset of AI that uses statistical techniques to enable machines to 'learn' from data without being explicitly programmed for every task. Within ML lies Deep Learning (DL), which utilizes multi-layered Artificial Neural Networks to solve highly complex problems like facial recognition. Think of it as a hierarchy: All Deep Learning is Machine Learning, and all Machine Learning is AI, but the reverse is not true.

How do machines actually learn? There are three primary paradigms: 1. Supervised Learning: The model is fed 'labeled' data. For every input $x$, we provide a target output $y$. The goal is to learn a mapping function $f(x) = y$. 2. Unsupervised Learning: The model receives data without labels. It looks for hidden patterns or structures, such as clustering similar customers together. 3. Reinforcement Learning (RL): An 'agent' learns by interacting with an environment. It receives rewards for good actions and penalties for bad ones, aiming to maximize the cumulative reward $G_t = \sum_{k=0}^{\infty} \gamma^k r_{t+k+1}$.

A model is only as good as its data. To build a reliable model, we split our dataset into two parts: the Training Set and the Testing Set (often an 80/20 split). The model learns patterns from the training set. We then use the testing set—which the model has never seen—to evaluate its performance. If a model performs perfectly on training data but fails on testing data, it has suffered from Overfitting. This means it 'memorized' the noise in the training data rather than learning the general underlying pattern.