- HW 01 (due Tu, Jan 22, 9pm)
peer assessment due by Th 9pm
Intro; basic foraging
- HW 02 (due Tu, Jan 29, 9pm)
Movement strategies
- HW 03 (due Tu, Feb 05, 9pm)
Braitenberg vehicles
- HW 04 (due Tu, Feb 12, 9pm)
Threshold logic
- HW 05 (due Tu, Feb 26, 9pm)
Action selection I
- HW 06 (due Tu, Mar 05, 9pm)
Action selection II
- HW 07 (due Tu, Mar 12, 9pm)
Spike-based coding
- HW 08 (due Tu, Apr 02, 9pm)
Associative learning -
- HW 09 (due Tu, Apr 09, 9pm)
Spatial navigation
- HW 10 (due Tu, Apr 16, 9pm)
Reinforcement learning
- HW 11 (due MON, Apr 22, 9pm)
Bot-challenge game design
- Final Proj (MON, Apr 29, 9pm)
Bot-challenge final
|
Homework
MCB 419 Homework 5
The first part of the assignment is to complete the exam1 programming exercise as
described here:
Exam1 NetLogo Specifications
Completing this part will be worth 80% of your hw05 grade.
There is no template. You need to write your own code from scratch.
Once that piece is working properly, the second part of the assignment is to add
UFO's that act as 'predators' rather than 'prey' for the rocket. Specifically:
- add 5 UFOs with speed controlled by a slider, similar to the rocket slider
- the UFOs should use a finite state machine (FSM) to 'hover' for 100 ticks and
then 'chase' the rocket for 100 ticks. This is coded for you in the following template:
hw05_ufo.nlogo.
- When a rocket and ufo collide, the rocket should lose 3 points and the ufo should be
moved to a new random location in the black region
- Add a switch called 'draw-trails?' and a button called 'clear trails' that control
whether or not the rocket leaves a trail on the screen.
After integrating the UFO's with your exam1 simulation, the screen should be similar to the one shown below.
Successfully integrating the UFO interactions is worth 10% of the grade.
Finally, the third part of the assignment is to make the rocket 'smarter' so that,
in addition to doing taxis towards stars, it also avoids being caught by the UFOs.
This part of the assignment is open-ended and counts for up to 10% of the grade.
The smarter your implementation the more points you can earn. Some ideas:
- you might add a pair of UFO sensors, in addition to the light sensors, so the rocket can avoid UFOs
- you might use a finite-state-machine (FSM) approach to control the rocket's behavior
- you might have the rocket 'hide' in the brown region where the UFOs can't reach it
- you can change the criteria for ending the game (fixed time, fixed points, etc.)
- you can make the task more difficult by increasing the ufo-speed relative to the rocket-speed; start easy and then make the task progressively harder
- you can include comments or explanations about your implementation on the INFO tab
Mail your completed netlogo file to mcb419@gmail.com with 'hw05' in the subject line.
Sample solution files:
hw05_sample_solution.nlogo
|