IIT M FOUNDATION AN EXAM QDF4 26 Oct - 26 Oct 2025

Concepts this paper tests

Auto-generated summary from the source site. Handy as a checklist; the lessons above explain each idea from scratch.

Here’s a structured summary of core topics, concepts, principles, and equations for the "Diploma Level: Machine Learning Foundations" exam, based on the provided question paper. The exam covers Machine Learning (ML) basics, Linear Algebra, and Mathematical Foundations. --- ---

1. Machine Learning Foundations

Core Topics & Concepts

1. Supervised Learning

   • Classification (Q2):

     • Predicts discrete labels (e.g., loan approval: "Yes/No").
     • Algorithms: Logistic Regression, Decision Trees, SVM, k-NN.
     • Key Idea: Map input features (e.g., income, credit score) to a class.

   • Regression:

     • Predicts continuous values (e.g., house price).
     • Algorithms: Linear Regression, Polynomial Regression.

   • Loss Functions (Q8):

     • Squared Loss (MSE): $L(y, \hat{y}) = (y - \hat{y})^2$.
     • Used in regression to measure prediction error.

1. Unsupervised Learning

   • Clustering (Q2 option C):

     • Groups similar data points (e.g., customer segmentation).
     • Algorithms: k-Means, Hierarchical Clustering.

   • Dimensionality Reduction (Q2 option D):

     • Reduces feature space (e.g., PCA, t-SNE).

1. Model Evaluation

   • Bias-Variance Tradeoff:

     • Bias: Error due to overly simplistic assumptions.
     • Variance: Error due to excessive sensitivity to training data.

   • Overfitting/Underfitting (Q9):

     • Overfitting: Model performs well on training data but poorly on test data (e.g., high-degree polynomial).
     • Underfitting: Model is too simple to capture patterns.

2. Linear Algebra

Core Topics & Equations

1. Matrix Basics

   • Row Space: Span of a matrix’s rows.

     • Basis (Q10): Linearly independent rows that span the row space.
     • Example: For matrix $A$, basis = $\{(1,0,2), (0,1,-1), (0,0,-1)\}$.
   • Rank: Number of linearly independent rows/columns.

   • Invertibility (Q12, Q14, Q16):

     • A square matrix $A$ is invertible iff $\det(A) \neq 0$.
     • Not all diagonalizable matrices are invertible (Q14: False).

1. Special Matrices

   • Orthogonal Matrix (Q12 option A):

     • $A^T A = I$ (columns are orthonormal).
     • Preserves vector lengths: $\|Ax\| = \|x\|$.

   • Symmetric Matrix (Q12 option C, Q13, Q15):

     • $A = A^T$.
     • Eigenvalues: Always real (Q15: True).
     • Diagonalizable: Always diagonalizable (even without distinct eigenvalues; Q16: False).

1. Eigenvalues & Diagonalization (Q13, Q15, Q16)

   • Eigenvalue Equation: $A v = \lambda v$.

   • Diagonalizable Matrix:

     • $A = P D P^{-1}$, where $D$ is diagonal.
     • Condition: Must have $n$ linearly independent eigenvectors (not necessarily distinct eigenvalues; Q16: False).

   • Real Eigenvalues:

     • Guaranteed for symmetric matrices (Q15: True).
     • Not guaranteed for non-symmetric real matrices (Q13: Option A).

1. Vector Spaces

   • Basis (Q10): Minimal spanning set for a subspace.
   • Linear Independence: No vector in the set can be written as a combination of others.

3. Mathematics (Calculus & Algebra)

Key Equations/Results

1. Derivatives & Optimization (Implied in Q8)

   • Squared loss gradient: $\frac{\partial}{\partial w} (y - \hat{y})^2 = -2(y - \hat{y})$.
   • Used in gradient descent for regression/classification.

1. Exponential/Logarithmic Identities (Q5, Q6, Q7, Q18)

   • $e^{\ln x} = x$, $\ln(e^x) = x$.
   • $3^4 \times 3^3 = 3^{4+3} = 3^7 = 2187$ (Q5 answer: 72 likely refers to $3^2 \times 2^3 = 9 \times 8 = 72$).
   • $2^{10} = 1024$ (Q6).
   • $\ln(1) = 0$ (Q7).
   • $\log_b(b) = 1$ (Q18).

2. Linear Algebra Concepts

3. Mathematics Dependencies

Key Takeaways for Exam Prep

1. ML Algorithms:

   • Know when to use classification vs. regression vs. clustering.
   • Understand loss functions (MSE) and evaluation metrics (overfitting).

1. Linear Algebra:

   • Basis/row space: Find linearly independent rows.
   • Eigenvalues: Symmetric matrices have real eigenvalues; diagonalization requires $n$ independent eigenvectors.
   • Orthogonal/symmetric matrices: Properties and examples.

1. Math:

   • Memorize exponent/log rules (e.g., $\log_b(b) = 1$).
   • Understand derivatives for optimization (e.g., MSE gradient).