BITS WILP AIML Knowledge Base: Semester 1 Resources

MFML is universally acknowledged as one of the toughest subjects in Semester 1. It requires genuine conceptual clarity in Linear Algebra, Multivariate Calculus, Matrix Decompositions, and Probability. This is not a subject you can cram the night before an exam; however, with structured concept practice, it can become your highest scoring subject.

ISM is widely regarded as one of the easiest and most scoring subjects in Semester 1. It covers measures of central tendency, probability distributions, hypothesis testing, ANOVA, and regression. However, because everyone performs well, scoring 85% might only net you a 'B' or 'C' grade due to relative grading. Solve practice problems from the J. L. Devore textbook to lock in an 'A'.

DNN is one of the most practically relevant subjects in the modern AI ecosystem. It covers perceptrons, activation functions, backpropagation derivations, optimization algorithms (Adam, RMSProp), convolutional networks (CNNs), and recurrent architectures (RNNs, LSTMs). Exams test a mix of numerical weight updates, architecture justifications, and reasoning.

Machine Learning covers supervised and unsupervised algorithms, decision trees (ID3, C4.5, entropy, information gain), naive Bayes classifiers, support vector machines (SVMs), ensemble methods, and clustering. Conceptual clarity and practicing calculations for decision tree splits will secure maximum marks in both EC2 and EC3.