Skip to content

ashbellett/ashbellett.github.io

Repository files navigation

Ash Bellett

This is my personal website that contains notes on mathematics, science and engineering for data scientists.

Prerequisites

Install Ruby.

Installation

Install Ruby gems: bundle install

Run website locally: bundle exec jekyll serve

Topics

1. Set theory

  • Logic
    • Logical statements
    • Neccesity, sufficiency and equivalence
    • Negation, conjunction and disjunction
  • Sets
    • Membership, equality and subsets
    • Predicates, quantifiers and specification
    • Set operations
    • Cardinality, finiteness and countability
  • Functions
    • Functions and composition
    • Injection, surjection and bijection
    • Images and inverses

2. Real analysis

  • Limits
    • Absolute value, distance and closeness
    • Sequences
    • Boundedness and monotonicity
    • Convergence and limits
  • Derivatives
    • Continuity and differentiability
    • Univariate differential calculus
    • Multivariate differential calculus
  • Integrals
    • Partitions and integration
    • Fundamental theorems of calculus
    • Univariate integral calculus
    • Multivariate integral calculus

3. Topology

  • Topological spaces
    • Topologies and neighbourhoods
    • Relative topologies and subspaces
    • Closure, continuity and compactness
  • Metric spaces
    • Metrics
    • Balls and points
    • Closure, open and closed sets
    • Boundedness and continuity
    • Connectedness and equivalance

4. Linear algebra

  • Vectors
    • Vector spaces
    • Linear independence
    • Span, basis and dimension
    • Inner product spaces
    • Positive definiteness
    • Length and angles
    • Orthogonality
  • Matrices
    • Matrix algebra
    • Determinants
    • Invertibility
    • Special matrices
  • Eigenvalues and eigenvectors
    • Eigen-decomposition
    • Characteristic polynomial
    • Diagonalisation
  • Matrix decompositions
    • Cholesky decomposition
    • QR decomposition
    • Singular value decomposition
  • Matrix applications
    • Systems of linear equations
    • Projection
    • Linear transformations

5. Probability

  • Probability spaces
    • Sample spaces and events
    • Sigma algebras
    • Probability measures
    • Probability axioms
  • Probability forms
    • Marginal probability
    • Joint probability
    • Conditional probability
  • Random variables
    • Pushforward probability measure
    • Support
    • Discrete random variables
    • Continuous random variables
    • Independent random variables
    • Transformations of random variables
  • Moments
    • Expectation
    • Variance
    • Higher-order moments
    • Covariance and correlation
    • Probability generating functions
    • Moment generating functions
    • Characteristic functions
  • Distributions
    • Discrete uniform
    • Bernoulli
    • Binomial
    • Geometric
    • Poisson
    • Continuous uniform
    • Exponential
    • Gamma
    • Beta
    • Chi-squared
    • Normal
    • Dirichlet
    • Multivariate distributions
  • Concentration bounds
    • Univariate bounds
    • Bounds of expectations
    • Bounds of sums
    • Bounds of functions
  • Probabilistic convergence
    • Pointwise and uniform convergence
    • Convergence in distribution
    • Convergence in probability
    • Almost-sure convergence
    • Delta method
  • Limit theorems
    • Laws of large numbers
    • Central limit theorem
    • Order statistics
  • Stochastic processes
    • Covariance and correlation
    • Stationarity
    • Gaussian processes
    • Wiener process
    • Renewal processes
    • Markov processes

6. Statistics

  • Samples
    • Summary statistics
    • Sufficiency
  • Estimators
    • Biasness
    • Consistency
    • Efficiency
  • Point estimation
    • Likelihood function
    • Method of least squares
    • Method of moments
    • Maximum likelihood estimation
  • Interval estimation
    • Interval estimators
    • Coverage probability
    • Confidence intervals
  • Hypothesis testing
    • Null and alternative hypotheses
    • Test statistic and critical value
    • Significance level and p-values
    • Power and sample size
    • Types of hypothesis tests
    • Simple and composite hypotheses
    • Multiple testing

7. Optimisation

  • Convexity
    • Convex sets and functions
    • Determination of convexity
    • Implications of convexity
  • Unconstrained optimisation
    • Gradient descent methods
    • Newton's method
  • Constrained optimisation
    • Substitution method
    • Lagrangian multipliers
    • Simplex method
    • Interior point methods

8. Supervised learning

  • Linear models
    • Linear regression
    • Logistic regression
    • Multinomial regression
    • Generalised linear models
  • Non-linear models
    • Polynomial regression
    • Spline regression
    • Fourier and wavelet bases
    • Generalised additive models
  • Kernel models
    • Maximal margin classifier
    • Support vector classifier
    • Support vector machines
  • Tree models and ensembling
    • Classification and regression trees
    • Bagging
    • Random forests
    • Boosting
    • Voting and stacked ensembles
  • Resampling methods
    • Cross-validation
    • Bootstrap methods
  • Feature selection
    • Exhaustive search
    • Subset selection
    • Recursive feature elimination
    • Regularisation penalties
  • Model selection
    • Adjusted R-squared
    • Mallows's Cp
    • AIC and BIC
    • Loss functions
    • Validation metrics

9. Unsupervised learning

  • Dimension reduction
    • Principal components analysis
    • Canonical correlation analysis
    • Factor analysis
    • Independent components analysis
  • Manifold learning
    • Multi-dimensional scaling
    • Isometric feature mapping
    • Local linear embeddings
    • Stochastic neighbourhood embeddings and t-SNE
    • Spectral embeddings
  • Density estimation
    • Gaussian mixture models
    • Expectation-maximisation algorithm
    • Histogram estimators
    • Kernel density estimators
  • Clustering
    • K-means clustering
    • K-medoids and PAM
    • Affinity propagation
    • Spectral clustering
    • Agglomerative hierarchical clustering
    • DBSCAN
    • Biclustering
    • Cluster evaluation
  • Novelty and outlier detection
    • One-class support vector machine
    • Elliptic envelope
    • Isolation forest
    • Local outlier factor
  • Association rule learning
    • Apriori algorithm
    • ECLAT algorithm
    • FP-growth algorithm
  • Covariance estimation
    • Empirical covariance
    • Shrinkage methods
    • Minimum covariance determinant

10. Bayesian methods

  • Subjective probability
    • Subjective uncertainty
    • Standard events
    • Conditional probability
    • Decisions
    • Utility
    • Estimation and prediction
  • Prior and likelihood representation
    • Exchangeability
    • De Finetti's representation theorem
    • Priors
    • Asymptotics
  • Parametric modelling
    • Conjugate models
    • Exponential families
    • Non-conjugate families
    • Posterior summaries
  • Computational inference
    • Intractable integrals
    • Monte Carlo estimation
    • Markov chain Monte Carlo (MCMC)
    • Hamiltonian Markov chain Monte Carlo
    • Analytic approximations
  • Model choice
    • Model uncertainty
    • Model averaging
    • Model selection
    • Posterior predictive checking
  • Linear models
    • Conjugate prior
    • Reference prior
    • General basis functions
    • Generalised linear models
  • Non-parameteric models
    • Random probability measures
    • Dirichlet processes
    • Pólya Trees
    • Partition models
    • Gaussian processes
    • Spline models
    • Partition regression models
  • Mixture models
    • Finite mixture models
    • Dirichlet process mixture models
    • Mixed-membership models
    • Latent factor models
  • Graphical models
    • Belief networks
    • Markov networks
    • Factor graphs

11. Deep learning

  • Activation functions
    • Linear
    • Sigmoid
    • RELU
    • ELU
    • Softmax
  • Loss functions
    • Mean squared error
    • Cross-entropy loss
    • Cosine similarity
    • KL divergence
  • Optimisers
    • Stochastic gradient descent
    • Momentum
    • Nesterov momentum
    • Adagrad
    • RMSprop
    • Adam
  • Initialisers
    • Glorot/Xavier initialisation
    • Orthogonal initialisation
  • Regularisation
    • Weight sharing
    • Dropout
    • Weight regularisation
    • Early stopping/patience
  • Normalisation
    • Batch normalisation
    • Layer normalisation
  • Feed-forward networks
    • Neuron unit
    • Multi-layer perceptron
  • Convolutional networks
    • Convolution operation
    • Pooling
    • Padding and stride
    • Transposed convolutions
  • Recurrent networks
    • RNN cell
    • Stacked and bi-directional RNNs
    • Back-propagation through time
    • Long short term memory cell
  • Transformer networks
    • Multi-head attention
    • Positional encoding
    • Transformer architecture
  • Normalising flows
    • Bijectors
    • Autoregressive flows
    • Masked autoregressive flow (MAF)
    • Masked autoencoder for distribution estimation (MADE)
    • Inverse autoregressive flow (IAF)
    • Non-linear independent components estimation (NICE)
    • RealNVP model
    • Glow model
  • Auto-encoder networks
    • Bottleneck architecture
    • Evidence lower bound (ELBO)
    • Reparameterisation trick
    • Variational auto-encoder
  • Bayesian networks
    • Aleatoric and epistemic uncertainty
    • Bayes by backpropogation
    • MC dropout
    • Uncertainty estimation

12. Agent systems

  • Multi-armed bandits
    • Optimal arm identification
    • Regret minimisation
    • Optimistic approaches
    • Thompson sampling
    • Contextual bandits
    • Bayesian optimisation
  • Reinforcement learning
    • Markov decision processes
    • Dynamic programming
    • Q-learning
    • Importance sampling
    • Linear function approximation
    • Deep Q-learning
    • Policy gradient methods

13. Structured data analysis

  • Data types
  • Data summaries
  • Data transformations
  • Data quality
  • Data visualisation
  • Temporal data
  • Spatial data
  • Multivariate data

14. Unstructured data analysis

  • Networks
    • Vertices and edges
    • Adjacency and direction
    • Neighbourhoods
    • Paths and cycles
    • Cliques and separation
    • Graph data structures
    • Graph embeddings
    • Search algorithms
    • Pathfinding algorithms
    • Minimum weight spanning tree
    • Community detection
    • Graph classification
  • Images
    • Image histograms
    • Affine transformations
    • Denoising
    • Edge detection
    • Feature detection
    • Image segmentation
    • Pose estimation
    • Motion estimation
    • Stereo correspondence
    • Object recognition
  • Language
    • Word classification
    • Word embeddings
    • Text clustering
    • Sequence modelling

15. Big data methods

  • Optimisation
    • Gradient descent methods
    • Distributed, stochastic gradient descent
    • Stochastic variational inference
  • Markov chain Monte Carlo
    • Divide and conquer methods
    • Subsampling methods
  • Streaming
    • Parameter estimation
    • Forgetting factors
    • Change point detection

Releases

No releases published

Packages

No packages published