JAWAHARLAL NEHRU TECHNOLOGICAL UNIVERSITY HYDERABAD B.Tech III MACHINE LEARNING Previous Question Paper (August - 2022 - R18)

Machine learning is an interdisciplinary field that draws foundational principles from several key areas:

1. Statistics: Provides the theoretical basis for learning from data, including concepts like probability distributions, hypothesis testing, and regression analysis. For example, linear regression models use statistical estimation to predict continuous values.
2. Computer Science: Focuses on the efficiency of algorithms and data structures. It addresses computational complexity and the design of architectures like neural networks. For example, the use of decision trees relies on graph theory and search algorithms.
3. Mathematics (Linear Algebra & Calculus): Essential for optimization and data representation. Linear algebra handles high-dimensional data via matrices, while calculus (specifically gradient descent) is used to minimize error functions.
4. Neuroscience: Inspired the development of artificial neural networks, which mimic the biological structure of neurons to perform pattern recognition, such as in image classification tasks.
5. Control Theory: Contributes to reinforcement learning, where an agent learns to make sequences of decisions to maximize a reward, similar to how a thermostat regulates temperature.

Machine learning models are primarily categorized based on how they learn from data:

1. Supervised Learning: The model is trained on a labeled dataset, where the input data is paired with the correct output. The algorithm learns a mapping function from input to output. Examples include classification (e.g., spam detection) and regression (e.g., house price prediction).
2. Unsupervised Learning: The model works with unlabeled data to discover hidden patterns or intrinsic structures. Common techniques include clustering (e.g., customer segmentation) and dimensionality reduction (e.g., Principal Component Analysis).
3. Semi-supervised Learning: A hybrid approach using a small amount of labeled data and a large amount of unlabeled data. This is useful when labeling is expensive or time-consuming, such as in medical image analysis.
4. Reinforcement Learning: An agent learns to make decisions by interacting with an environment. It receives feedback in the form of rewards or penalties, aiming to maximize cumulative reward over time. This is widely used in robotics, game playing, and autonomous navigation.