ToDo-List & Ideas

Roadmap

Here are some ideas we hope to cover/implement in the future.

Time: Thursday 1:00 pm CDT

Location: Loomis 322

Current schedule (2019 Spring)

DatePersonSubject
Jan. 17PaulKriging
Jan. 31RyanRandomized SVD
Feb. 14EliSpectral Embedding and Clustering
Feb. 28ChrisMarkov Decision Process and Reinforcement Learning
Mar. 14MattThe Art of O(1) Lookup
Mar. 28BrianB-tree
May 23Will WeiAdaboost
May 30AlinaNatural Language Generation

2016 schedule

2017sp schedule

2017fa schedule

2018sp schedule

Speaker Guidelines

TL;DR Do unto others as you would have them do unto you.

At minimum, an AIG presenter should prepare a few slides and an example code.

  1. The slides should explain: - what the algorithm is - how a minimum example works - why the algorithm might be practically useful
  2. A code that demonstrates the simplest problem the algorithm solves.
  3. Make a pull-request (PR) to the website repository to make your presentation and code eternal. Follow instructions in the README.md file.

The presentation should be 30-50 min if given without interruptions. Interactive elements are encouraged. e.g. a Jupyter notebook demo. with tweakable parameters given by the audience.

Ideas for presentations

  • Machine learning.
    • Back propogation.
    • Clustering.
    • Boltzmann machine.
  • Control theory and signal processing.
    • Model reduction.
    • Kalman filter.
    • Hidden markov model.
    • Proportional-integral (PI) controller.
  • Stochastic algorithms.
    • The Metropolis approach to sampling and alternatives
    • Global balance (pentalty method).
    • Quasi-random numbers.
    • Ant Colony Optimization
    • Parallel tempering.
    • Stochastic hill climbing.
    • Bayesian networks.
    • Simulated annealing
    • Evolutionary, and genetics algorithms.
    • Particle Swarm Optimization
    • Belief propogation
    • Gibbs sampling
  • Encryption.
    • Symmetric-key, Public-key (RSA) cryptography
    • Cryptanalysis (breaking encryption).
    • Hashing.
  • Optimzation.
    • Global Newton methods
      • line search
      • trust region
      • iterative solution of linear equations
      • matrix free
    • Generalized minimal residual method (GMRES)
      • preconditioning
      • additive Schwarts
      • Algebraic and geometric multi-grid
      • Block Jacobi
    • Quadratic optimization.
    • Convex optmization.
    • Steepest descent, Conjugate gradiant, Quasi-Newton, ….
    • Noisy optimization.
    • Compilers (fortan).
    • Simplex method.
  • Linear Algebra.
    • Random matrix theory.
    • QR / SVD. principle component analysis
    • Diagnolization, inversion.
    • Lanczos
    • Fast Fourier transforms (FFT).
  • Numerical solutions to differential equations.
    • Finite differences
    • Finite elements
    • Finite volumes
    • Spectral elements
    • PDE solvers (additional problems from multivariate)
    • Energy conserving or time-reversal invariant versions.
    • Runge-Kutta and family.
  • Data compression.
    • Image compression techniques (one or more).
    • Compressed sensing (probabilistic approach and connections to stat mech).
    • Compressed sensing (l-1 technique).
  • Image Processing
    • Image recognition
    • Automatic focus
  • Visualization
    • Marching Cubes
  • Quantum computing.
    • Quantum annealing.
    • Quantum error correction.
    • Quantum encryption.
    • Quantum stabilizers.
    • Grover algorithm.
    • Shor algorithm.
  • Parallelism.
    • Parallel linear algebra.
    • OpenMP and MPI
    • GPU, Cuda, …
  • Computer networks/the internet.
    • Google search bar, page rank.
    • The Internet protocal suite.
    • Packet switching vs. cell-based switching.
    • Mobile networks.
    • Error detection and correction, Hamming codes.
    • Internet security.
    • Network routing.
  • Classic CS algorithms
    • Quicksort, Graph theory, …
    • Cellular atomata.
    • Theory of computation (Turing, finite state machine, definition of language, regular expressions).
    • complexity theory.