Master Data Analytics

An interactive, single-page masterclass. Navigate through the premium handwritten notes, mathematical proofs, code templates, and solved assignments for SSMDA (Units 3 & 4).

Syllabus Overview

UNIT III

Data Analytics Foundations

Describe classes of open and closed set. Apply the concept of compactness. Describe Metric space - Metric in Rn. Use the concept of Cauchy sequence, completeness, compactness and connectedness to solve problems.

UNIT IV

Advanced Concepts in Data Analytics

Describe vector space, subspaces, independence of vectors, basis and dimension. Describe Eigen values, Eigen vectors and related results.

Unit 3: Data Analytics & Topology

1. Metric Spaces (X, d)

Let X be any non-empty set. A metric space is a pair (X, d) where d: X × X → [0, ∞) is a distance function (metric) that satisfies four essential axioms for any x, y, z ∈ X:

1. Non-negativity: d(x,y) ≥ 0
2. Identity: d(x,y) = 0 ⇔ x = y
3. Symmetry: d(x,y) = d(y,x)
4. Triangle Inequality: d(x,y) ≤ d(x,z) + d(z,y)

Important Examples of Metric Spaces:

Usual Metric on ℝ: d(x,y) = |x - y|. This is the standard distance on the real number line.
Discrete Metric: d(x,y) = 0 if x=y, and 1 if x≠y. In this space, every point is isolated.
Euclidean Metric in ℝ²: d(x,y) = √((x₁-y₁)² + (x₂-y₂)²). This is the straight-line distance in 2D space.
Pseudo-metric Space: Similar to a metric space, but d(x,y) can be 0 even if x ≠ y. For example, d(x,y) = |x² - y²|. Here d(1,-1) = 0, but 1 ≠ -1. Hence, it is a pseudo-metric, not a metric.

2. Spheres, Open Sets & Closed Sets

Before defining open sets, we must define an Open Sphere (or Ball). Let p ∈ X and r > 0. The open sphere S(p,r) is the set of all points x ∈ X such that d(x,p) < r.

Open Sphere: S(p,r) = {x ∈ X : d(x,p) < r}
Closed Sphere: S[p,r] = {x ∈ X : d(x,p) ≤ r}

Open Set: A subset U ⊂ X is open if, for every point x ∈ U, there exists an open sphere (ε-neighborhood) centered at x that is entirely contained within U.
Example: The open interval (a,b) in ℝ is an open set.

Closed Set: A set F ⊂ X is closed if its mathematical complement (X \ F) is an open set.
Example: The closed interval [a,b] in ℝ is closed because its complement (-∞, a) ∪ (b, ∞) is open.

Key Properties:

Arbitrary union of open sets is open; Finite intersection of open sets is open.
Arbitrary intersection of closed sets is closed; Finite union of closed sets is closed.
The Real Line (ℝ) and the Empty Set (∅) are both clopen (open AND closed simultaneously).

Limit Points & Closure: A point 'p' is a limit point of set 'S' if every neighborhood of 'p' contains an infinite number of members of 'S'. The Closure of a set ($\overline{A}$) is the union of the set A and its derived set (set of all limit points A'). So, $\overline{A} = A \cup A'$.

3. Covers & Compactness

An Open Cover of a set S is a collection of open sets whose union completely contains S. A Subcover is a smaller selection from that cover that still manages to contain S.

Compact Space: A topological space (or subset) is compact if every open cover of the set has a finite subcover. Essentially, you can always reduce an infinite number of covering sets down to a finite number.

Heine-Borel Theorem:
Every closed & bounded subset of real numbers (ℝ) is compact.
- [0, 1] is compact (it is closed and bounded).
- (0, 1) is NOT compact (bounded but not closed).
- ℝ is NOT compact (closed but not bounded).

4. Cauchy Sequences & Completeness

A sequence {xₙ} in a metric space converges to 'x' if for a given ε > 0, there exists an integer m such that d(xₙ, x) < ε for all n ≥ m. Every convergent sequence has a unique limit.

Cauchy Sequence: A sequence where the terms get arbitrarily close to each other as the sequence progresses.

∀ ε > 0, ∃ integer p such that:
d(xₙ, xₘ) < ε for all n, m ≥ p

Theorem: Every convergent sequence is a Cauchy sequence. However, the converse is not always true!

Complete Metric Space: A metric space is said to be COMPLETE if every Cauchy sequence in that space converges to a limit that is also within that space. For example, ℝ is a complete metric space. The set of rational numbers ℚ is incomplete (because a Cauchy sequence can approach √2, which is not in ℚ).

5. Connected vs. Disconnected Sets

Two non-empty subsets A and B are Separated if $A \cap \overline{B} = \emptyset$ and $\overline{A} \cap B = \emptyset$. Note: separated sets are disjoint, but disjoint sets aren't always separated.

Disconnected Set: A subset S is disconnected if it can be expressed as the union of two non-empty separated sets (S = A ∪ B).

Connected Set: A set that is not disconnected. In the real line ℝ, a set is connected if and only if it is an interval or a singleton point.

Examples: S = (0,1) ∪ (2,3) is disconnected. S = [0,1] is connected.

Unit 4: Advanced Concepts (Linear Algebra)

1. Rank of a Matrix & Vector Spaces

Rank of a Matrix: The highest order of a non-zero minor of a matrix. In simpler terms, it is the number of linearly independent rows or columns in a matrix that cannot be made completely zero using elementary matrix operations.

Vector Space V(F): Let V be a set of elements called Vectors, and F be a field of elements called Scalars. The set V is a Vector Space over F if it satisfies the following 10 axioms:

Internal Composition (V, +) is an Abelian Group:

Closure Axiom: For all α, β ∈ V, α+β ∈ V.
Associativity: (α+β)+γ = α+(β+γ).
Existence of Identity (Zero Vector): α + 0 = α.
Existence of Inverse: For every α, there is -α such that α + (-α) = 0.
Commutativity: α + β = β + α.

External Composition (Scalar Multiplication): V is closed under scalar multiplication. For all a ∈ F and α ∈ V, a·α ∈ V.

Distributive & Associative Scalar Axioms:

a·(α + β) = a·α + a·β
(a + b)·α = a·α + b·α
a·(b·α) = (ab)·α
1·α = α (where 1 is the multiplicative identity in F)

2. Vector Subspaces

If W is a subset of V(F) and W itself forms a vector space over the same field F using the same operations, then W is a Subspace of V(F).

Subspace Theorem (One-Step Test):
A non-empty subset W of a vector space V(F) is a subspace if and only if:
For all a, b ∈ F and for all α, β ∈ W ⟹ aα + bβ ∈ W

Important Property: The intersection of two subspaces (W₁ ∩ W₂) is always a vector subspace. Proof: Since both contain the zero vector, intersection is non-empty. For any a,b ∈ F and α,β ∈ W₁ ∩ W₂, the linear combination aα+bβ belongs to W₁ (since it's a subspace) and also belongs to W₂, thus aα+bβ ∈ W₁ ∩ W₂.

3. Linear Span & Independence

Linear Combination: A vector α is a linear combination of vectors α₁, α₂... αₙ if there exist scalars a₁, a₂... aₙ such that:
α = a₁α₁ + a₂α₂ + ... + aₙαₙ

Linear Span (L(S)): The set of all possible linear combinations of a set of vectors S. It is denoted by L(S) and is always a subspace of V(F).

Linear Dependence vs. Independence: Let S = {α₁, α₂... αₙ}. Consider the equation: a₁α₁ + a₂α₂ + ... + aₙαₙ = 0.

If this equation implies that a₁ = a₂ = ... = aₙ = 0 (only the trivial solution exists), the vectors are Linearly Independent.
If at least one scalar aᵢ ≠ 0, the vectors are Linearly Dependent. This means at least one vector can be expressed as a linear combination of the others.

4. Basis and Dimension

A subset S of V(F) is called a Basis of V(F) if it satisfies two strict rules:

S is a set of Linearly Independent vectors.
L(S) = V(F). That is, every element of V(F) can be generated as a linear combination of elements in S (S spans the space).

Dimension: If V(F) has a finite basis, the number of elements in the basis is called the Dimension of V(F), denoted as dim(V). For example, the standard basis for ℝ³ is {(1,0,0), (0,1,0), (0,0,1)}, hence dim(ℝ³) = 3.

Theorem: Every finite-dimensional vector space has at least one basis, and any two bases of the same space will always have the exact same number of elements.

5. Eigenvalues and Eigenvectors

Let A be an n×n square matrix. A scalar λ is an Eigenvalue (or Characteristic Value) of A if there exists a non-zero column vector X such that:

AX = λX ⟹ (A - λI)X = 0

The vector X is the Eigenvector corresponding to λ.

Working Steps to find them:

Create the Characteristic Equation by taking the determinant: |A - λI| = 0.
Solve the resulting polynomial equation to find the roots. These roots are the Eigenvalues (λ₁, λ₂, ...).
For each eigenvalue λᵢ, substitute it back into the equation (A - λᵢI)X = 0.
Solve this homogeneous system of linear equations to find the vector X. This is the Eigenvector. There are infinite eigenvectors for a single eigenvalue (they form an eigenspace), but we usually find the simplest non-zero base vector.

Solved Assignments

Assignment No. 3 (DA-304T)

Q1. What is the meaning of data analytics?

Data Analytics is the comprehensive science of analyzing raw datasets to extract meaningful insights, draw conclusions, and identify actionable trends. In the modern computational era, it involves applying algorithmic, statistical, and mechanical processes to solve complex problems.

For a 5-mark scope, it generally encompasses four key types:

Descriptive Analytics: Examines what has happened in the past (e.g., monthly revenue reports).
Diagnostic Analytics: Investigates why something happened (e.g., finding the cause of a sudden drop in website traffic).
Predictive Analytics: Uses historical data and ML algorithms to forecast future outcomes.
Prescriptive Analytics: Recommends specific actions to achieve desired outcomes.

Overall, its core motive is to optimize business performance, drive evidence-based decision-making, and reduce operational risks.

Q2. Explain properties of closed & open set.

Open Sets: A set $ U $ is called open if, for every point $ x \in U $, there exists an $\epsilon$-neighborhood entirely contained within $ U $. It does not include its boundary points.

The arbitrary union of any number of open sets is always an open set.
The finite intersection of open sets is an open set.
Example: The open interval $ (0,1) $ in $ \mathbb{R} $.

Closed Sets: A set $ F $ is closed if its complement $ X \setminus F $ is an open set. Alternatively, it contains all its limit points and boundary points.

The arbitrary intersection of closed sets is always a closed set.
The finite union of closed sets is a closed set.
Example: The closed interval $ [0,1] $ in $ \mathbb{R} $.

Q3. Explain compactness and compact space.

Compactness is a fundamental topological property that generalizes the concept of a closed and bounded subset in Euclidean space to more abstract spaces.

Formal Definition: A topological space (or subset) $ K $ is said to be compact if every open cover of $ K $ has a finite subcover. This means if you cover the set with infinitely many open sets, you can always pick a finite number of those sets that still completely cover $ K $.

Heine-Borel Theorem: In Euclidean space $ \mathbb{R}^n $, a subset is compact if and only if it is both closed and bounded.

Examples:

The closed interval $ [0,1] $ is compact because it is bounded (length of 1) and closed (includes its limits).
The open interval $ (0,1) $ is not compact; it is bounded but not closed.
The set of real numbers $ \mathbb{R} $ is closed, but it is not bounded, hence not compact.

Q4. What is metric space. With example.

A Metric Space is an ordered pair $ (X, d) $ where $ X $ is a non-empty set and $ d $ is a metric (distance function) mapping $ X \times X \to [0, \infty) $. To qualify as a metric space, the function $ d $ must satisfy four rigid axioms for any elements $ x, y, z \in X $:

Non-negativity: $ d(x,y) \ge 0 $
Identity of indiscernibles: $ d(x,y) = 0 \iff x = y $
Symmetry: $ d(x,y) = d(y,x) $
Triangle Inequality: $ d(x,z) \le d(x,y) + d(y,z) $

Examples:

Real Line with Usual Metric: $ (\mathbb{R}, d) $ where $ d(x,y) = |x - y| $.
Euclidean Metric Space: $ \mathbb{R}^n $ where the distance between two points $ x $ and $ y $ is $ \sqrt{\sum (x_i - y_i)^2} $.

Q5. Describe concept of Cauchy sequence.

A Cauchy Sequence is a sequence in a metric space where the elements become arbitrarily close to each other as the sequence progresses. It evaluates the "closeness" of terms to one another rather than to a final limit.

Formal Definition: A sequence $ \{x_n\} $ in a metric space $ (X, d) $ is Cauchy if, for every arbitrarily small positive real number $ \epsilon > 0 $, there exists a positive integer $ N $ such that for all indices $ m, n > N $, the distance $ d(x_m, x_n) < \epsilon $.

Significance & Completeness:

Every convergent sequence is a Cauchy sequence.
However, the reverse is not always true. A Cauchy sequence might not converge if the space has "holes" (e.g., a sequence of rational numbers converging to $ \sqrt{2} $ does not converge within the set of rational numbers $ \mathbb{Q} $).
If every Cauchy sequence in a metric space converges to a point within that space, the space is called a Complete Metric Space.

Assignment No. 4 (DA-304T)

Q1. Define vector space and its motive.

Definition: A Vector Space $ V(F) $ is an algebraic structure consisting of a set of elements called vectors, defined over a field of scalars $ F $. It must satisfy two primary operations: vector addition and scalar multiplication.

To qualify as a vector space, $ V $ must be an Abelian group under addition (satisfying closure, commutativity, associativity, additive identity, and additive inverse), and it must satisfy scalar multiplication axioms (distributivity over scalar and vector addition, scalar associativity, and scalar identity).

Motive/Purpose:

It abstracts the geometric concept of physical vectors (like forces and velocities) into higher dimensions ($ \mathbb{R}^n $).
It forms the foundational framework for Linear Algebra, enabling the systematic solving of systems of linear equations.
In Data Analytics and ML, data sets are represented as high-dimensional vector spaces, allowing for operations like dimensionality reduction (PCA) and vector distance calculations.

Q2. Internal and external composition.

In abstract algebra and vector spaces, operations are categorized based on the sets from which they draw their operands.

1. Internal Composition: A binary operation that takes two elements from the same set and produces a resultant element that also belongs to that same set. This represents closure.
Example in Vector Spaces: Vector Addition. If you take vector $ \alpha $ and vector $ \beta $ from vector space $ V $, their sum $ \alpha + \beta $ must also belong to $ V $.

2. External Composition: A binary operation that combines an element from one set (e.g., a Field of scalars) with an element from a different set (e.g., a Vector space) to produce an element in the vector space.
Example in Vector Spaces: Scalar Multiplication. If you take a scalar $ a \in F $ and a vector $ \alpha \in V $, their product $ a \cdot \alpha $ must belong to the vector space $ V $.

Q3. Basis & dimensions of subspace W of ℝ³.

(i) $ W = \{(a,b,c) : a+b+c=0\} $:
From the given condition, we can express $ c $ in terms of free variables $ a $ and $ b $: $ c = -a-b $.
Substitute this into the general vector $ (a,b,c) $:
$ (a, b, -a-b) = (a, 0, -a) + (0, b, -b) $
$ = a(1,0,-1) + b(0,1,-1) $.
The vectors $ (1,0,-1) $ and $ (0,1,-1) $ span the subspace $ W $ and are linearly independent.
Basis: $ \{(1,0,-1), (0,1,-1)\} $
Dimension: 2

(ii) $ W = \{(a,b,c) : a=b=c\} $:
Let the common value be a scalar parameter $ k $. Then any vector in $ W $ takes the form:
$ (k, k, k) = k(1, 1, 1) $.
The single non-zero vector $ (1,1,1) $ spans the entire subspace $ W $ and forms a linearly independent set.
Basis: $ \{(1,1,1)\} $
Dimension: 1

Q4. Eigen value and eigen vectors of matrix A.

Given Matrix A =

113 151 311

1. Characteristic Equation: We set the determinant $ |A - \lambda I| = 0 $.
Expanding the determinant yields the cubic equation: $ \lambda^3 - 7\lambda^2 + 36 = 0 $.

2. Eigenvalues ($\lambda$): Solving the cubic equation gives roots:
$ (\lambda + 2)(\lambda - 3)(\lambda - 6) = 0 $.
Thus, Eigenvalues are $ \lambda_1 = -2, \lambda_2 = 3, \lambda_3 = 6 $.

3. Eigenvectors: Substitute each $\lambda$ into $ (A - \lambda I)X = 0 $:
- For $ \lambda = -2 $: Solve $ (A + 2I)X = 0 $. Row reduction yields $ X_1 = [1, 0, -1]^T $.
- For $ \lambda = 3 $: Solve $ (A - 3I)X = 0 $. Row reduction yields $ X_2 = [1, -1, 1]^T $.
- For $ \lambda = 6 $: Solve $ (A - 6I)X = 0 $. Row reduction yields $ X_3 = [1, 2, 1]^T $.

Q5. Write two pros and cons of Data Analytics.

Pros of Data Analytics:

Evidence-Based Decision Making: It removes guesswork and intuition from business strategies, allowing executives to make decisions based on concrete historical patterns and predictive models, leading to higher success rates.
Operational Efficiency & Optimization: By analyzing supply chain, workflow, and customer data, organizations can identify bottlenecks, reduce waste, optimize marketing budgets, and significantly improve their ROI.

Cons of Data Analytics:

Data Privacy and Security Risks: Gathering and processing massive amounts of consumer data makes organizations highly susceptible to data breaches, cyber-attacks, and severe legal penalties if privacy laws (like GDPR) are violated.
High Complexity and Implementation Costs: Establishing a robust analytics pipeline requires expensive infrastructure (cloud databases, software) and highly skilled, highly paid personnel (Data Scientists, Engineers). Poor data quality can also lead to misleading conclusions.