Home    Course Archive    Video/Audio Library     Papers Archive    CV    My Keeper Shelf    Cultural References in Class Quotes

smh: COSC 681 Simulation
FALL 2016


0. Basics
  1. Anaconda Cheat Sheet
  2. cosc681.wikispaces.com
  3. Required Textbook:


  4. Data sets: A Students Guide to Python

  5. Development:

  6. Syllabus

  7. Coding Standards (Keep it clean. Keep it readable. Keep it honest. Keep it simple.)
    Points will be subtracted for bad readability (use of white space, choice of variable names) as well as for lack of functionality and poor design.

  8. Office Hours: See instructor home page



1. Lectures
Date Topic Reading Assignments / Due
12/7 More on HMM: watch the video on HMM https://www.youtube.com/watch?v=jY2E6ExLxaw
12/5 Hidden Markov Model Joint probability distribution formula: HiddenMarkovModel.doc
  • https://webcourse.cs.technion.ac.il/236522/Spring2008/ho/WCFiles/tutorial05.ppt
  • http://setosa.io/blog/2014/07/26/markov-chains/
  • https://www.youtube.com/watch?v=TPRoLreU9lA
  • https://www.youtube.com/watch?v=jY2E6ExLxaw
11/30 Markov chain (1) https://www.analyticsvidhya.com/blog/2014/07/markov-chain-simplified/
(2) https://www.analyticsvidhya.com/blog/2014/07/solve-business-case-simple-markov-chain/
(3) https://www.analyticsvidhya.com/blog/2014/07/interpret-hidden-state-latent-markov-model/
(4) https://www.dartmouth.edu/~chance/teaching_aids/books_articles/probability_book/Chapter11.pdf
Monte Carlo: Math examples
11/2 Your models
Euler forward method (e.g., p 105, Ean 6.26)
Introduction to Runge-Kutta Methods
Numerical ODE methods (Euler's, Heun's, midpoint)
Runge Kutta 2nd order method: Basics
Runge-Kutta (4th order) Method Introduction
Runge Kutta 4th order Method

Matrix exponential
Matrix Exponential 1
Matrix Exponential 2
See below
10/31 7.1, 7.2
10/26 6.1-6.4 See below
10/24 5.6, 5.7 Exercise the Fibonacci system (Eqn 5.48) and visualize its phase space as in Exercis 5.11.
Do Exercise 5.12 and Exercise 5.13. Demo 5.12 and 5.13 on Monday 10/31!
10/19 Cobweb graph
10/12 Modelling Complex Systems - intro
outline
Sayama: Chpts 1, 2, 3, 4 Exercise the Python in chapter 4 (nothing to turn in)
10/10 Initializing the wiki: cosc681.wikispaces.com
9/26 AnyLogic
from MIT University Course
Overview of Agent-Based Modeling Part 1
Overview of AB Modeling Part 2
Introdution to Interface and supporting concepts
Specifying Agent Properties, Parameters & Heterogeneity
Discrete Intra-Agent Dynamics, Statecharts
Discrete Inter-agent Dynamic, Messaging
Events, Static & Dynamic

Course videos by Prof Johan Joubert --basic assembly line
basic assembly part 1
basic assembly part 2
basic assembly part 3
Pick up and drop off Building state transitions:
pickup and drop off
Joubert: Custom resource unit (state transition)
9/12 Petri net
(1)http://www.ulb.ac.be/di/ssd/ggeeraer/Tutorial-Perti-Nets-Geeraerts.pdf
(2) http://www.isr.umd.edu/Labs/CIM/miscs/wmsor97.pdf
(3)https://www.researchgate.net/publication/3217035_Petri_nets_and_industrial_applications_A_tutorial
(4)Petri Net Modelling and Simulation in AnyLogic and MATLAB for ARGESIM Benchmark C4 'Dining Philosophers'
(1) Read researchgate Petri net paper
(2) Read AnyLogic and Petri net sne-journal paper
(3) Exercise AnyLogic: HW0912 Due 9/21 9/28
9/7 Orientation Follow along the tutorial for Agent-Based modelling example in AnyLogic 7 in Three Days



2. Homework

Item Date
Hidden Markov Model, part 2: HW 12/07
Help with PART A, Solution to find P(X1) given no observations
More help with PART A, #1. You need P( X(1) | Y1=c ) to find P( X(2) | Y(1:1) ), i.e., using observation at previous time
Now to solve P(X2 | Y1=c, Y2=c), first predict P(X2 | Y1) using filtered P(X1 | Y1), then filter P(X2 | Y2)
PART B, Find P( X(t) | Y(1:t) ) -- See the recurrence given approx 48:00. You use both recurrences, filtering and prediction.
12/12
Hidden Markov Model: HW 12/05 12/12
Monte Carlo: HW 11/28 12/12
Various from Sayama: HW 11/14 11/14
Sayama: Exercise 6.5: Compare errors using Euler;s forward and a 4th order RungeKutta method
Exercise (6.7 or 6.8): Implement using Euler's forward method and using 4th order Runge Kutta method. Compare errors
11/9
Sayama: Exercise 5.12 and 5.13 ,Exercise 6.5 and (6.7 or 6.8) 10/26, Demo on 11/2
AnyLogic FAQs 9/30

3. Test

Last changed: