Beskjeder
Suggested solutions for the exam can be found here. Thanks for following the course, have a nice summer!
By popular demand we have made a quick solution to last years exam, see the directory previous exams. Let me know if you find any bugs :)
See for Q&A at 1215 in auditorium 2
Martin
There will be a Q&A session wednesday june 5. at 1215 in Vilhelm Bjerknes' hus Auditorium 2 (NB).
I have posted last years exam-problems in previous exams.
I hope to see many of you there :)
Solveig Bruvoll and Martin Syre Wiig are researchers at the Norwegian Defence Research Establishment (FFI). They will present their research regarding path planning methods for autonomous vehicles. The path planning is based on optimization methods like Dijkstra and A* for calculating the shortest path through a graph. The aim is to develop a path planning method that produces situation dependent paths. They will explain the optimization method that is used and how the method is applied to their applications.
On tuesday we will make a brief summary of the curriculum and then go through the exam for 2017
I encourage everyone to register for the Abel lectures (free lunch included) Wednesday, May 22, see
http://www.abelprize2019.com
On tuesday we went through the primal and dual simplex method applied to network problems, and thus almost finished the chapter V14 (which I encourage everyone to read, together with Geir Dahls lecture notes). The next lecture will be the final ordinary lecture. We will finish section V14 and discuss some applications of network LPs, such as the famous Dijkstra algorithm.
In the tutorial of tomorrow (10 April) I will not go through the correction of any exercize. Instead, I will use the tutorial to clarify any possible issue or question about the second mandatory assignment.
... the slides are posted on the webpage. We finished convex analysis (Geir Dahls lecture notes). Good luck with the oblig, and have a nice easter - see you for Network Flow Problems april 23rd!
In this lecture we will implement the simplex algorithm on MATLAB. The algorithm is based on the theory covered in the lecture of Feb 19, so it is useful to look again at these slides before Tuesday. Also, you are encouraged to bring your own laptop at the lecture.
The second mandatory assignment is available.
Here you can download the pdf file with the text.
Additionally, to complete the last exercise, you will need to download the file data.mat (or, alternatively, the following three files: A.txt, b.txt, c.txt)
The deadline is Thursday 11 April.
Important: to have a passing grade you must have satisfactory answers to at least 50% of the questions and have seriously attempted to solve ALL OF THEM.
I have posted the lecture notes from yesterdays lecture.
Referring to discussions in class, it seems like some of you struggle with the double optimization in the minmax and maxmin problems. It may be useful to think of the result of the inner-optimization (the max in the minmax problem) as a function (only) of the variable in the outer optimization. Hence the outer variable is fixed in the inner optimization (only the inner variable varies). Note also that the result of the inner optimization is the objective function of the outer optimization (Linear-Program)
Todays lecture at 1015-1200 is cancelled due to illness. Sorry for the short notice
The results of the first mandatory assignment are now available on Devilry. A new deadline for the second and last attempt has been set on Tue 12 March.
Today we finished chapter 6 (matrix formulations), did chapter 7.1 (sensitivity) and started on section 17 (interior point methods), see the posted lecture-notes
Note: You should read Chapters 6.3-6.5 (example, dual matrix simplex and twostep methods) yourselves (self-study/homework).
Next week we will continue in chapter 17 (interior point methods), first very briefly refreshing Lagrange-multiplier theory.
We made it halfway through "Duality" (chapter 5) yesterday - see the annotated/revised lecture-notes on the course-pages (under "Lectures"). To be continued next week.
As I mentioned there are a number of readable explanations/motivations for duality out there, check out this one for instance
http://www.science4all.org/article/duality-in-linear-programming/
The first mandatory assignment is available here.
The deadline is Thursday 28 February.
Important: to have a passing grade you must have satisfactory answers to at least 50% of the questions and have seriously attempted to solve ALL OF THEM.
The student representative is Eline Larsen Valen (elinelv [alfakr?ll] student.matnat.uio.no). Any comments or suggestions regarding the course (e.g., lectures, exercises, book, etc.) can be communicated to Eline.
Some of you may find it easier to execute OPL projects from the command line. Here you find a simple guide explaining how to do that.
Don't forget about the weekly assignments! They are posted weekly on the Timeplan page.
Go to this website.
Find the label “Get student and faculty editions for free” and click on it.
Click “Register” (top-right corner).
Type your institutional e-mail address under “Organization-Issued Email Address”
Go on with the registration (UiO is among the recognized universities, so the registration should work smoothly).
Once you are registered, sign in.
Type “CPLEX” in the search bar (“Product search”).
Choose the first result (“IBM ILOG CPLEX Optimization Studio v12.8 (CJ2IKML)”), and click “Add to cart”.
Download the software (you can choose among Windows, Linux and Mac versions) and install it normally.