# MA304: Topics in Applied Mathematics

Unit 1: Linear Analysis   In Unit 1, the theory of linear algebra studied in the Saylor Foundation’s MA211 and MA212 are extended to linear analysis in that matrices are extended to linear operators that include certain differential operators.  In this unit, you will study the inner product and its corresponding norm defined on a vector space, along with their important properties that depend on the Cauchy-Schwarz inequality.  In addition, you will review the eigenvalue problem, and you will study singular values with an application to spectral decomposition.  This leads to the discussion of singular value decomposition (SVD) of rectangular matrices that allows us to generalize the inversion of nonsingular matrices, studied in the Saylor course MA211, to the “inversion” of rectangular and singular square matrices with applications to solving arbitrary systems of linear equations and to the introduction of the method of principal component analysis (PCA).  As an application of PCA, the formulation and theory for data dimensionality reduction (DDR) will also be studied in this first unit.

Unit1 Learning Outcomes
Upon successful completion of this unit, the student will be able to:
- Compute inner products of two vectors, which include infinite sequences and functions, and compute the corresponding norms. - Apply the Cauchy-Schwarz inequality to find the angle between two vectors. - Compute eigenvalue and eigenvector pairs. - Compute singular values of arbitrary rectangular matrices. - Perform SVD. - Find first and second principle components of data matrices. - Explain the essence of data dimensionality reduction. - Reduce the dimension of data matrices for simple examples.

1.1 Inner Product and Norm Measurements   1.1.1 Definition of Inner Product   - Reading: Cambridge University Press: Marcus Pivato’s Linear Partial Differential Equations and Fourier Theory: “6A: Some Functional Analysis: Inner Products” Link: Cambridge University Press: Marcus Pivato’s Linear Partial Differential Equations and Fourier Theory:“6A: Some Functional Analysis: Inner Products”(PDF)

`````` Instructions: Please click on the link above to access the PDF, and
study Section 6A on pages 103–105, stopping at Section 6B, to learn
Studying this reading should take approximately 15 minutes to
complete.

displayed on the webpage above.
``````
• Reading: University of California, Davis: Isaiah Lankham, Bruno Nachtergaele, and Anne Schilling’s Linear Algebra: As an Introduction to Abstract Mathematics Link: University of California, Davis: Isaiah Lankham, Bruno Nachtergaele, and Anne Schilling’s Linear Algebra: As an Introduction to Abstract Mathematics (PDF)

Instructions: Please click on the link above, and then select the link “PDF version of the book” to download the text.  Study Section 9.1 on pages 117–119 for a definition and examples of inner product.  You will be using this text throughout the course, so you may find it helpful to save the PDF to your desktop.
Studying this reading should take approximately 15 minutes to complete.

`````` Instructions: Please click on the link above, and view the
derivation of the Cauchy Schwarz inequality for the Euclidean
space.
Viewing the lecture and pausing to take notes should take
approximately 30 minutes to complete.

displayed on the webpage above.
``````
• Reading: Northern Illinois University: John A. Beachy’s “Theorem 5.3: Cauchy-Schwarz Inequality” and Wikipedia’s “Cauchy-Schwarz Inequality” Links: Northern Illinois University: John A. Beachy’s “Theorem 5.3: Cauchy-Schwarz Inequality” (HTML) and Wikipedia’s “Cauchy-Schwarz Inequality” (HTML)

Instructions: Please click on the links above, and read these webpages in their entirety to study the proof of the Cauchy-Schwarz inequality for the general inner-product space.  Please note that these readings also apply to the topics outlined in sub-subunits 1.1.3 and 1.1.4.
Studying the proofs in the reading materials takes approximately 30 minutes to complete.

1.1.3 Norm Measurement and Angle between Vectors   - Reading: University of California, Davis: Isaiah Lankham, Bruno Nachtergaele, and Anne Schilling’s Linear Algebra: As an Introduction to Abstract Mathematics Link: University of California, Davis: Isaiah Lankham, Bruno Nachtergaele, and Anne Schilling’s Linear Algebra: As an Introduction to Abstract Mathematics (PDF)

`````` Instructions: Please click on the link titled “PDF version of the
book” to access the text.  Study Sections 9.3 through 9.6 on pages
119–135 for information on the general theory and properties of the
inner product and its associated norm.
Studying this text should take approximately 1 hour and 15 minutes
to complete.

displayed on the webpage above.
``````

1.1.4 Gram-Schmidt Orthogonalization Process   - Lecture: MIT: Professor Gilbert Strang’s Linear Algebra: “Lecture 17: Orthogonal Matrices and Gram-Schmidt” Link: MIT: Professor Gilbert Strang’s Linear Algebra: “Lecture 17: Orthogonal Matrices and Gram-Schmidt” (YouTube)

`````` Instructions: Please click on the link above, and view this entire
lecture to learn about orthogonal matrices, orthonormal families,
and the Gram-Schmidt procedure for finding an orthonormal family
from a given linearly independent family.
Viewing this lecture and pausing to take notes should take
approximately 1 hour to complete.

displayed on the webpage above.
``````

1.2 Eigenvalue Problems   1.2.1 Linear Transformations   - Lecture: MIT: Professor Gilbert Strang’s Linear Algebra: “Lecture 30: Linear Transformations and Their Matrices” Link: MIT: Professor Gilbert Strang’s Linear Algebra: “Lecture 30: Linear Transformations and Their Matrices” (YouTube)

Instructions: Please click on the link above, and view the entire lecture on linear transformations.
Viewing this lecture and pausing to take notes and understanding the lecture should take approximately 1 hour to complete.

`````` Terms of Use: Please respect the copyright and terms of use
displayed on the webpage above.
``````

1.2.2 Bounded Linear Functionals and Operators   - Reading: Bounded Linear Functionals and Operators The Saylor Foundation does not yet have materials for this portion of the course. If you are interested in contributing your content to fill this gap or aware of a resource that could be used here, please submit it here.

``````[Submit Materials](/contribute/)
``````

1.2.3 Eigenvalues and Eigenspaces   - Lecture: MIT: Professor Gilbert Strang’s Computational Science and Engineering I: “Lecture 6: Eigen Values (Part 2) and Positive Definite (Part 1)” Link: MIT: Professor Gilbert Strang’s Computational Science and Engineering I: “Lecture 6: Eigen Values (Part 2) and Positive Definite (Part 1)” (YouTube)

`````` Instructions: Please click on the link above, and view the entire
Viewing this lecture and pausing to take notes should take
approximately 1 hour and 15 minutes to complete.

displayed on the webpage above.
``````
• Reading: University of California, Davis: Isaiah Lankham, Bruno Nachtergaele, and Anne Schilling’s Linear Algebra: As an Introduction to Abstract Mathematics: “Section 7: Eigenvalues and Eigenvectors” Link: University of California, Davis: Isaiah Lankham, Bruno Nachtergaele, and Anne Schilling’s Linear Algebra: As an Introduction to Abstract Mathematics (PDF)

Studying this reading should take approximately 15–20 minutes to complete.

1.2.4 Self-Adjoint Positive Definite Operators   - Lecture: MIT: Professor Gilbert Strang’s Linear Algebra: “Lecture 27: Positive Definite Matrices” Link: MIT: Professor Gilbert Strang’s Linear Algebra: “Lecture 27: Positive Definite Matrices” (YouTube)

`````` Instructions: Please click on the link above, and view the entire
lecture on positive definite matrices.
Viewing this lecture and pausing to take notes should take
approximately 1 hour and 15 minutes to complete.

displayed on the webpage above.
``````

1.3 Singular Value Decomposition (SVD)   - Lecture: MIT: Gilbert Strang’s Linear Algebra: “Lecture 29: Singular Value Decomposition” Link: MIT: Gilbert Strang’s Linear Algebra: “Lecture 29: Singular Value Decomposition” (YouTube)

`````` Instructions: Please click on the link above, and view the entire
that this video lecture also covers the topics outlined in
sub-subunits 1.3.1 through 1.3.3.
Viewing this lecture and pausing to take notes should take
approximately 1 hour to complete.

displayed on the webpage above.
``````

1.3.1 Normal Operators and Spectral Decomposition   - Reading: University of California, Davis: Isaiah Lankham, Bruno Nachtergaele, and Anne Schilling’s Linear Algebra: As an Introduction to Abstract Mathematics Link: University of California, Davis: Isaiah Lankham, Bruno Nachtergaele, and Anne Schilling’s Linear Algebra: As an Introduction to Abstract Mathematics (PDF)

Studying this reading should take approximately 1 hour to complete.

`````` Terms of Use: Please respect the copyright and terms of use
displayed on the webpage above.
``````

1.3.2 Singular Values   Note: This topic is covered by the reading assigned below sub-subunit 1.3.1

1.3.3 Reduced Singular Value Decomposition   Note: This topic is partially covered by the reading assigned below sub-subunit 1.3.1.

• Reading: Reduced Singular Value Decomposition The Saylor Foundation does not yet have materials for this portion of the course. If you are interested in contributing your content to fill this gap or aware of a resource that could be used here, please submit it here.

Submit Materials

1.3.4 Full Singular Value Decomposition   - Reading: Full Singular Value Decomposition The Saylor Foundation does not yet have materials for this portion of the course. If you are interested in contributing your content to fill this gap or aware of a resource that could be used here, please submit it here.

``````[Submit Materials](/contribute/)
``````

1.4 Principal Component Analysis (PCA)   1.4.1 Frobenius Norm Measurement   - Reading: Frobenius Norm Measurement The Saylor Foundation does not yet have materials for this portion of the course. If you are interested in contributing your content to fill this gap or aware of a resource that could be used here, please submit it here.

``````[Submit Materials](/contribute/)
``````

1.4.2 Principal Components for Data-Dependent Basis   - Reading: Principal Components for Data-Dependent Basis The Saylor Foundation does not yet have materials for this portion of the course. If you are interested in contributing your content to fill this gap or aware of a resource that could be used here, please submit it here.

``````[Submit Materials](/contribute/)
``````

1.4.3 Pseudoinverses   - Lecture: MIT: Professor Gilbert Strang’s Linear Algebra: “Lecture 33: Left and Right Inverses: Pseudoinverse” Link: MIT: Professor Gilbert Strang’s Linear Algebra: “Lecture 33: Left and Right Inverses: Pseudoinverse” (YouTube)

`````` Instructions: Please click on the link above, and view the entire
lecture to learn about the topic of matrix pseudoinverses and its
application to least-squares estimation.
Viewing this lecture, pausing to take notes, and studying the
material in the lecture should take approximately 3 hours to
complete.

displayed on the webpage above.
``````

1.4.4 Minimum-Norm Least-Squares Estimation   - Reading: Minimum-Norm Least-Squares Estimation The Saylor Foundation does not yet have materials for this portion of the course. If you are interested in contributing your content to fill this gap or aware of a resource that could be used here, please submit it here.

``````[Submit Materials](/contribute/)
``````

1.5 Application to Data Dimensionality Reduction   - Reading: Application to Data Dimensionality Reduction The Saylor Foundation does not yet have materials for this portion of the course. If you are interested in contributing your content to fill this gap or aware of a resource that could be used here, please submit it here.

``````[Submit Materials](/contribute/)
``````

1.5.1 Representation of Matrices by Sum of Norm-1 Matrices   1.5.2 Approximation by Matrices of Lower Ranks   1.5.3 Motivation to Data-Dimensionality Reduction   1.5.4 Principal Components as Basis for Dimension-Reduced Data