Sandbox

PageTree

What is Artificial Intelligence

Course — Life Cycle

Classroom 77 hours
Lab 33 hours
Project 90 days

Course Details

  1. Supervised Learning
    1. Generative/Discriminative learning
    2. Parametric/Non Parametric learning
    3. Neural Network
    4. Support Vector Machines
  2. Unsupervised Learning
    1. Clustering
    2. Dimensionality reduction
    3. Kernel Methods
  3. Learning Theory
    1. Bias/Variance Tradeoffs
    2. VC Theory
    3. Large Margins
  4. Reinforcement Learning
    1. Adaptive Control

Books

  1. Christopher Bishop, Pattern Recognition and Machine Learning
  2. Richard Duda, Peter Hart and David Stork, Pattern Classification
  3. Tom Mitchell, Machine Learning
  4. Richard Sutton and Andrew Barto, Reinforcement Learning

Session Details – Prerequisites – 1 hour 15 minutes each

  1. Linear Algebra - 15 Classes
    1. Basic concepts of notations
    2. Matrix Vector Multiplication
    3. Operations and Properties
      1. Transpose, Symmetric
      2. Trace, Norms
      3. Independence, Rank, Inverse
    4. Matrix Calculus
      1. The gradient
      2. The Hessian
      3. Eigen Values as Optimization
  2. Probability Theory – 10 Classes
    1. Elements of Probability
    2. Random Variables
    3. Cumulative Distribution Functions
    4. Probability Mass Functions
    5. Probability Density Functions
    6. Expectations
    7. Variance
    8. Common Random Variables (Bernoulli, Binomial)
    9. Joint and Marginal Functions
    10. Multiple Random Variables
  3. Convex Optimization Overview —10 Classes
    1. Lagrange Duality
    2. The Lagrangian
    3. Primal and Dual Problems
    4. Interpreting the primal and dual problems
    5. Complementary Slackness
    6. The KKT Conditions
    7. The L1 Norm Soft Margin SVM

Session Details – Machine Learning – 1 hour 15 minutes each

  1. Supervised Learning – 15 Classes
    1. Logistic Regression, Perceptron, Exponential Family
    2. Generative Learning Algorithm, Gaussian Discriminant Analysis, Naïve Bayes
    3. Support Vector Machines
    4. Model Selection and Feature Selection
    5. Ensemble Methods: Bagging, Boosting, ECOC
    6. Evaluating and Debugging Learning Algorithms
  2. Learning Theory – 7 Classes
    1. Bias and Variance Tradeoff Union and Chernoff/Hoeffding bounds
    2. VC Dimension. Worst Case (online) Learning
    3. Advice on how to use learning Algorithms
  3. Unsupervised Learning – 11 Classes
    1. Clustering and K-means
    2. EM, Mixture of Gaussians
    3. Factor Analysis
    4. PCA, MDS, pPCA
    5. Independent Component Analysis (ICA)
  4. Reinforcement Learning and Control - 9 Classes
    1. MDPs. Bellman Equations
    2. Value iteration and Policy iteration
    3. Linear Quadratic regulation (LQR).LQG
    4. Q-Learning. Value function approximation
    5. Policy Search. Reinforce. POMDPs

Projects

  1. Recommending partners in social networking platforms as per one’s behavioral characteristics and Needs
  2. Railways Departure Delay Prediction
  3. Good Reads Recommendations
  4. Finding common Opinions in User-Generated Reviews
  5. Grouping Photos by Face
  6. Internet Article Comment Classifiers
  7. Train your TV

Projects

  1. Corporate Valuation
  2. Detecting Corporate Frauds
  3. Are you hot or not?
  4. Emotion detection from Speech

Living in Unity

Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License