Skip to content

Learning notes on ML and practice projects, covering key areas such as computer vision (handwritten digit recognition), boosting techniques, data preparation, data visualization, e2e pipelines, with practical applications including RAi predictions for ATP, UFC, and EPL.

Notifications You must be signed in to change notification settings

dianamatkava/ml-ds-projects

Repository files navigation

Machine Learning projects and notes

Minesweeper Game

Vectors Properties

  • [[Vectors]]:
    • Mathematical objects representing quantities with both magnitude and direction, often expressed in coordinate form.
  • [[Linear combinations and spans]]
    • A linear combination involves summing scaled versions of vectors;
    • span is the set of all possible linear combinations of a given set of vectors.
  • [[Programmimg/Math/Subspaces and the basis for a subspace|Subspaces and the basis for a subspace]]
    • A subspace is a set of vectors closed under vector addition and scalar multiplication.
    • A basis is a minimal set of vectors that span the subspace.
  • [[Linear Dependance and Independence]]
    • Linear Independence: No vector can be expressed as a combination of others. (unique solution exists). There is not noise.
    • Linear Dependence: At least one vector can be represented as a combination of others. Infinitely many solutions or no solution may exist.

Manual methods to solve system of linear equations:

  • Crammer Method (2x2) (3x3):
    • A method using determinants to solve systems of linear equations. It works well for small systems but is inefficient for larger ones. If the determinant == 0, the system has no unique solution.
  • [[Gaussian Elimination]]:
    • A systematic method for simplifying systems of linear equations through row operations to reach an upper triangular form, facilitating back substitution.
  • [[Gauss-Jordan Elimination]]:
    • An extension of Gaussian elimination that transforms the matrix to reduced row echelon form (RREF) ( results to [[I - Identity Matrix]]), allowing for direct reading of solutions without back substitution.

Matrix Properties

  • [[Matrix Vector Multiplication]]
    • linear transformation by multiplication.
  • [[Basis for the Column Space of a Matrix]]
    • A set of linearly independent column vectors in a matrix that spans the column space, crucial for understanding the matrix's range.
  • [[Inverse Matrix]]inverse is just like identity divided by matrix
    • $A \cdot A^{-1} = A^{-1} \cdot A = I$
    • Adjugate and Determinant Method:
      • (2x2): $A^{-1} = \frac{1}{ad - bc} \begin{pmatrix} d & -b \ -c & a \end{pmatrix}$
      • (>3x3): $\text{det}(E) = a(ei - fh) - b(di - fg) + c(dh - eg)$
    • [[Gauss-Jordan Elimination]]:
      • Augmented Matrix Transformation: $[U ∣ I]$ into the form $[I ∣ U^{-1}]$
      • Why it works? : let say $U=2$, $I=1$, what is $U^{-1}$?
        • $2 | 1$ => $\frac{R1}{2}$ => $\frac{2}{2}$ => $1|0.5$
        • $2*0.5=1$

About

Learning notes on ML and practice projects, covering key areas such as computer vision (handwritten digit recognition), boosting techniques, data preparation, data visualization, e2e pipelines, with practical applications including RAi predictions for ATP, UFC, and EPL.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published