Introduction to Agent-based Modeling

I took this course primarily because I had enjoyed precious offerings from Complexity Explorer. I had very little programming background, and my expectations for this course were mixed, given that, at first sight, NetLogo looks a little clunky with 1990's style graphics. I also had the impression that ABM was a "soft" discipline.

I couldn't have been more wrong. I quickly came to appreciate that the point of ABM is the insights it delivers about otherwise intractable problems not the quality of the graphics. Bill Rand brings this out in a well-paced, almost gentle way. However before I knew it, I was immersed in the stuff. By the end of the course I felt that I knew Bill quite well. This is an illusion but it speaks to his friendly delivery.

To get the full benefit of this course, I would not recommend auditing or skimming. This is one MOOC where the real benefit comes from actually producing NetLogo models. If you don't have a programming background, you may well need to allow more time than is suggested. It's not as easy as Bill makes it look.

Finally, the quality of a MOOC is heavily impacted by the liveliness of the forums and the ABM forum was well supported by both learners and instructors

All round I was very pleasantly surprised, and heartily commend the course

The lecturer is an expert in agent-based modelling which is very valuable for this course. Personally I do not think I learnt a lot from it though, and that's partly because the course covers to a large fraction NetLogo and its features. A good part of this does not help you to understand Agent-based Modeling, but only NetLogo better. This might be fine to some or many students, but I think it should be made more clear in the course description.

Other than that there were a few other downsides: the course content could have a partial redesign in my opinion as some things are introduced in an odd order. In general, the target audience doesn't seem to be clearly defined. As an example, there is a unit with data analysis in R. For a beginner with low to zero programming skills it was very fast; for someone with programmatic data analysis skills it was too slow or even simply unnecessary. Lastly, I found it difficult sometimes to follow the videos because of the presentation style, as there were a lot of filler words for example.

All in all I think this course has potential, especially due to the experience of the lecturer, but it can be improved.

This course is the poorest programming course i ever attend!
1. In general, the course is all like reading out loud the textbook.
2. Short and insufficient programming lectures.
3. 70% is not about simulation in Netlogo at all.
4. Poor teaching, the instructor fails to deliver the course content, poor educative language and skills.
5. Lack of self-evaluation, exercises and proper quizzes. you will not found out that where you are, at which levels you are, how much you actually learn. The instructor introduce a project but it's peer review who actually may not learned properly to judge.
no feedback at all.
6. The course has advertise oriented content about Netlogo.
7. Long course (9 week) without rich content.
I suggest if anyone want to learn Netlogo, there is some exercises in "introduction to Complexity" which beautifully addressed use of Netlogo in the context. I actually learn Netlogo there more than this course!
Do not purchase!

This course is a great disappointment for those who want to learn agent-based modeling:
1. It is necessary to have some previous knowledge in programming and modeling.
2. The instructor has very serious teaching problems, he just reads the slides, which are a summary of the book.
3. Lack of adequate pedagogy and didactics. For example, the evaluation is based on punishment, confusion and arbitrary criteria.
4. The evaluations are truculent, delocalized, subjective and do not correspond to the information provided on the slides or in the "explanation" of the instructor. I saw myself failing on questions to which I knew the answer but which responded to the arbitrary criteria of the instructors. For example, in one of the evaluations the correct answer was one of these options: modify, recombine, mutate or delete. Recombining something is not modifying it? Is it not mutating some kind of evolutionary modification? The "correct" answer was: recombine.
5. Peer evaluation is arbitrary, subjective and inappropriate, students are evaluated as if they had prior knowledge in modeling.

For me this course was an important step forward in learning modeling after finding out about NetLogo in 'Model Thinking' course by Scott Page. Coming from the field of medicine and being passionate about modeling for several decades I did not expect that adding programming skills to this would be easy. But this course by Professor Rand @ Complexity explorer made me to change my attitude. Thus, the most attracting and valuable aspect of the course which guided my decision to enroll was the possibility to build a program for scratch. And it was indeed true. Of course it would be naive to consider that after this course I will become a skilled programmer, but it can serve as a very comprehensive road map to farther exploration of this field and effective leverage for creating new models. I would specially mention instructor's friendly and professional delivery. And finally, I really enjoyed the course and would recommend it to anyone with interest in ABM.

Very nice course. The topics of the lessons were meaningful and complete. You don't need previous experience in modeling nor in programming for taking this course. The first weeks give you an easy-to-underst and overview of the different types of models, why agent based models are useful and where can you apply them. The later weeks go deeper and features advanced abm techniques.

The lessons are example-based and are very concerete, so you can make use of the knowledge right away. I didn't give this couse 5 stars because the platform is yet inmature (has some bugs), and the proffesors seem to have no previous experience with online courses, thus the videos could be improved (for example, there are too much filler words). This might be a little annoying but should not prevent you from taking full advantage of the course.

Coming from a background of no previous experience with Agent-Based Modeling in general (and Netlogo in particular), this course is a great introduction: accessible and easy to follow, the core of this course will take you through the basic theoretical concepts to building your first model.

Obviously, being a novice, there were certain times where some subjects felt a bit overwhelming but the forum and office hours were extremely helpful and made sure that I was never too lost.

The only less positive aspects I have to mention were some of the units (particularly the one that relates to the use of R) feeling a bit rushed, and a couple of problems with the course site (particularly its peer-review system) that were eventually fixed.

In general, I find the course to be very informative and interesting. It is quite easy to pass (the grading system is designed to be this way), and I would classify the difficulty of the material as being akin to that of an introductory course in college. I enrolled in the course with quite a bit of background knowledge in different disciplines which probably helped a lot, but I would still say that the instructors do a good job in keeping the instructional material accessible to a large range of learners.

My only complaint is about the NetLogo aspect of the course. It's certainly a great tool for ABM, but I find that the NetLogo training was neither very comprehensive given the requirements for the final assignment (hence the 4 stars) nor particularly necessary for future ABM practitioners (just my personal opinion). The program is a bit niche -- designed for grad students and researchers -- and the intellectual buy-in for NetLogo in this course has the potential to edge out time that could be used to develop a concrete conceptual understanding of ABM. Novice coders will find NetLogo to be particularly unforgiving.

Great course and subject matter. The cost and payment system is very reasonable. I chose to pay slightly more than the minimum fee for the course, but still feel I got amazing value for money. Thank you to Complexity Explorer and its financial supporters for making this high quality course available all around the world at such a generous rate. Bill Rand is awesome and really knows his stuff (he co-wrote the definitive introductory textbook with Uri Wilensky, author of Netlogo). The course moves at a pretty comfortable pace, though complete beginners may find some later sections require closer attention/revision (so pay attention!). We got to design our own working model which was very satisfying and a great way to jump straight in to programming with Netlogo. On the whole I really enjoyed it and would recommend it to anyone with an interest in ABM.