JAWAHARLAL NEHRU TECHNOLOGICAL UNIVERSITY HYDERABAD B.Tech III MACHINE LEARNING Previous Question Paper (February - 2023 - R18)

Machine learning is a subfield of artificial intelligence that focuses on the development of algorithms and statistical models that enable computer systems to improve their performance on a specific task through experience. Rather than being explicitly programmed for every possible scenario, these systems identify patterns in data to make predictions or decisions.

Inductive bias refers to the set of assumptions that a machine learning algorithm uses to predict outputs for inputs it has not encountered during training. Since a learner must generalize from a finite set of training data to an infinite set of possible instances, it cannot rely solely on the data provided; it must incorporate prior knowledge or constraints to choose between multiple hypotheses that are consistent with the training data.

Key aspects include:

• It defines the search space of the learning algorithm, effectively narrowing down the possible functions the model can learn.
• Without an inductive bias, a model would simply memorize the training set (overfitting) and fail to generalize to new, unseen data.
• Examples include the assumption of linearity in linear regression, the preference for simpler models (Occam's razor), or the assumption of smoothness in neural networks.