Coding, JavaScript, Machine Learning, Milestones, Murderous Maths, neural networks, Notes on everyday life, Set the beautiful mind free, Simon teaching, Simon's Own Code, Simon's sketch book

What Simon did instead of taking the SAT on Saturday

On Saturday morning, Simon didn’t go to the SAT examination location, although we had registered him to try taking the SAT (with great difficulties, because he is so young). In the course of the past few weeks, after trying a couple of practice SAT tests on the Khan Academy website, we have discovered that the test doesn’t reveal the depth of Simon’s mathematical talent (the tasks don’t touch the fields he is mostly busy with, like trigonometry, topology or calculus and require that instead, he solves much more primitive problems in a strictly timed fashion, while Simon prefers taking time to explore more complex projects). This is what happens with most standardized tests: Simon does have the knowledge but not the speed (because he hasn’t been training these narrow skills for hours on end as his older peers at school have). Nor does he have the desire to play the game (get that grade, guess the answers he deosn’t know), he doesn’t see the point. What did he do instead on his Saturday? He had a good night sleep (instead of having to show up at the remote SAT location at 8 a.m.) and then he…

built an A.I. applying a genetic algorithm, a neural network controlling cars moving on a highway! The cars use rays to avoid the walls of the highway. Implementing neuroevolution. What better illustration does one need to juxtapose true achievement and what today’s school system often wants us to view as achievemnt – getting a high grade on a test? The former is a beautiful page from Simon’s portfolio, showing what he really genuinely can do, a real life skill, something he is passionately motivated to explore deeper, without seeking a reward, his altruist contribution to the world, if you will. The latter says no more than how well one has been trained to apply certain strategies, in a competitive setting.

Simon’s code is online: https://repl.it/@simontiger/Raytracing-AI

Simon has put this version on GitHub: https://github.com/simon-tiger/Raycasting-A.I.

He has also created an improved version with an improved fitness function. “In the improved version, there’s a feature that only shows the best car (and you can toggle that feature on and off). And most importantly, I am now casting relative to where it’s going (so the linearity is gone, but it jiggles a lot, so I might linear interpolate it)”, – Simon explains. You can play with the improved version here: https://repl.it/@simontiger/Raycasting-AI-Improved

Finally, Simon is currently working on a version that combines all the three versions: the original, the improved and the version with relative directions (work in progress): https://repl.it/@simontiger/Raytracing-AI-Full

“I am eventually going to make a version of this using TensorFlow.js because with the toy library I’m using now it’s surprisingly linear. I’m going to put more hidden layers in the network”.

The raytracing part of the code largely comes from Daniel Shiffman.

Simon’s two other videos about this project, that was fully completed in one day:

Part 1
Part 2


Biology, Coding, Java, JavaScript, Milestones, Simon makes gamez, Simon's Own Code

Simon has created an “immortal” organism?

The organism is the green triangle on the left

Simon opened up a genetic algorithm game he built about two years ago and made a fascinating discovery: one of the organisms seems to have become immortal! Simon has called his discovery “The Everlasting Vehicle” and saved the vehicle’s DNA.

Links to the game on GiHub:
Original code: https://github.com/simon-tiger/steering-behaviors-evolution
p5.js version: https://simon-tiger.github.io/Game_SteeringBehaviorsEvolution/SteeringBehaviours_EvolutionGame_p5/

Simon writes:
The last time I ran the program is a couple of hours ago. Everything died out, except for one vehicle.

Stats
I have programmed this with a genetic algorithm. They have a DNA with 4 genes.

Attraction/Repulsion to food
Attraction/Repulsion to poison
How far it can see food
How far it can see poison
They also have a health, which goes down over time. If they eat food, then their health goes up, if they eat poison, then their health suddenly goes down. A good health is 1, and a bad one is 0.

So what was The Everlasting Vehicle’s DNA and health?

Property Value
Attraction/Repulsion to food 1.9958444373034823
Attraction/Repulsion to poison 1.3554737395594456
How far it can see food 53.31017416626768
How far it can see poison 23.33902221893798
Average health ~397
So it attracts to poison, yet its health is approximately 397 times bigger than a very good health! And better yet, it even lasted for a couple of hours so far!!!

Credits:
Inspired by Daniel Shiffman’s Evolutionary Steering Behaviors Coding Challenge
Link to the Challenge: https://www.youtube.com/watch?v=flxOkx0yLrY

Biology, Coding, Java, live stream, Living Code, Milestones, Python, Simon makes gamez, Simon teaching

The Genetic Algorithm Game Bug Solved!

Simon’s version of the Citius Invaders game in Processing is finally working! The big bug that seemed so unsolvable (enemies endlessly crossing over) has been defeated!

Siraj Raval presented this game in Python during Week 9 of his “Math of Intelligence” course. Simon’s is a simplified version that still incorporates a genetic algorithm.

Link to code on github at: https://github.com/simon-tiger/citius-invaders/

Link to our previous post on this project: https://antwerpenhomeschooling.wordpress.com/2017/11/23/problem-with-a-genetic-algorithm-game/

Link to Simon’s archived Live Stream about the project: https://antwerpenhomeschooling.wordpress.com/2017/12/01/live-stream-30-november-speechjs-and-citius-invaders/

Coding, JavaScript, live stream, Milestones, Simon makes gamez, Simon teaching

Live Stream 30 November. Speechjs and Citius Invaders.

Below are the archived versions of Simon’s live streams today. The second part (starting with a genetic algorithm game in Processing/ p5.js) was an emotional roller-coaster, with the OBS software failing to capture sound and giving Simon a hard time jumping between scenes, and yet he didn’t give up.

Simon writes: In the first part of this live session, I redid my speechsynth.js tutorial (so it has some coding in it). In the second part (which will likely be WAY longer), I will do a Processing/p5 challenge about Citius Invaders.

 

Biology, Coding, Java, live stream, Milestones, Simon makes gamez

Problem with a Genetic Algorithm Game

Simon has been working very hard at making his version of the Citius Invaders game that Siraj Raval presented in Python during Week 9 of his “Math of Intelligence” course. It’s a simplified version that still incorporates a genetic algorithm. Simon’s plan is to code the same game in P5 during his next live tutorial (at 17:00 CET on 30 November). Unfortunately, after many lines of code in several files, when the game was nearly ready, he faced a problem he doesn’t know how to resolve. He tried getting some online help via the Coding Train Slack channel but didn’t get much feedback. The code is now on github at: https://github.com/simon-tiger/citius-invaders/

Here is how Simon describes the problem:

Hi! I’m working on a “Citius Invaders” game, and I have a problem.
In the game, there are enemies, a spaceship, and bullets that it shoots.
The parameters of the game are:
* The game starts with 6 enemies.
* The minimal amount of enemies is 4.
* The maximal amount of enemies is 100.
The rules of the game are:
* When a bullet collides with an enemy, the bullet and the enemy get deleted.
* If there are 4 enemies on the screen, you’re not allowed to shoot any bullets any more.
* If there are 100 enemies on the screen, you’ve lost the game.
* The enemies reproduce every 5 seconds (this amount of time decreases every generation).
The goal of the game is to stay as close to 4 enemies as possible.
The problem is that: After the 5 seconds, the enemies don’t stop crossing over! I think the order that I’m crossing them over is wrong. They breed (cross over) forever and that makes the game pause forever. Maybe the algorithm in which I let them cross over is wrong. I have a `while` loop somewhere, maybe it became a forever loop. The `while` loop is at line 52 in `Population.pde`:

 while (!selectedEnemies1.isEmpty() || !selectedEnemies2.isEmpty()) {
 int index1 = int(random(selectedEnemies1.size()));
 int index2 = int(random(selectedEnemies2.size()));
 if (index1 == index2) {
 index2 = (index2 + 1) % selectedEnemies2.size();
 }

Below are a some videos showing Simon at different stages of the project. In the top video, he speaks of the difficulties he encountered with the frame rate and the genetic algorithm.

 

 

Biology, Coding, Java, JavaScript

Error with Genetic Algorithm. What is wrong?

Simon was almost done translating Smart Rockets example no. 2 (Smart Rockets Superbasic) from Daniel Shiffman’s The Nature of Code, from Processing (Java) into JavaScript in Cloud9, when he got an error using genetic algorithm: the dna seems to be undefined while Simon did define mom and dad dna.

This is Simon’s translation online in Cloud9: https://ide.c9.io/simontiger/smart-rockets#openfile-README.md

This is when he first discovered the bug and tried different solutions:

And this is the same project before he introduced the genetic algorithm:

In the next video Simon boasts he found two errors in his code and hopes that the problem would be solved, but alas, the rockets still vanish from the canvas after a few seconds:

Simon is officially stuck here.

On the positive side, this project did get us to read more about the actual human DNA and the way it works.

 

Biology, Coding, Java, JavaScript, Murderous Maths, Physics

Smart Rockets Coding Challenge translated into Processing (Java)

And here is Simon’s translation of Daniel Shiffman’s Smart Rockets Coding Challenge into Processing (Java). This challenge (originally in JavaScript) was about implementing a genetic algorithm from scratch and create “smart rockets” (based on a simulation by Jer Thorp). The rockets evolve the “best” path to a target, moving around obstacles.

Coding, Java, JavaScript, Milestones, Murderous Maths, Simon's Own Code

Simon translated Travelling Salesperson Coding Challenge from JavaScript into Processing (Java)

Travelling Salesperson 3 May 2017

On Tuesday and Wednesday, Simon ventured upon a serious quest – he translated Daniel Shiffman’s multi-part Traveling Salesperson Coding Challenge from JavaScript into Processing (Java).

The travelling salesman problem (TSP) asks the following question: If you have a list of cities and you absolutely have to visit every city on the list exactly once, what is the shortest possible route you can take? It’s a classical problem on combinatorial optimization and is widely studied in theoretical computer science (where, given a length L, the task is to decide whether the graph has any tour shorter than L). The TSP has many applications in planning, manufacture of microchips, astronomy and even in DNA sequencing (where cities are DNA fragments).

Simon and I have read about this problem in a Dutch children’s book (Telduivel) where it was stated that if a salesperson has to visit 25 cities and thus there are 25! possible routes to take, no modern computer can handle calculating the shortest one. Simon decided to check what the maximum number of characters was for a long in Java, and it turned out to be 19 (while 25! equals to 15.511.210.043.330.985.984.000.000 and exceeds that maximum). In his TSP example, Daniel Shiffman uses only 10 cities and 10! equals to merely 3.628.800, so that worked out fine.

Wikipedia says that even though the problem is computationally difficult, a large number of algorithms are known, so that some instances with tens of thousands of cities can be solved completely and even problems with millions of cities can be approximated within a small fraction of 1%.[1]

Simon noticed that even with his modest number of 10 cities, Processing (Java) ran faster than p5.js (JavaScript).

In the first two videos he starts and explains the project, in the later videos he applies additional algorithms to enhance accuracy.

 

Lexicographic Order Algorithm, one algorithm to iterate over all the permutations of an array:

 

Genetic Algorithm:

 

 

In the following video, Simon explains what crossover is and how he plans to implement it within the Genetic Algorithm file in order to get even more exact results. (He later got stuck when introducing crossover into his Processing code).

Biology, Coding, Java, JavaScript, Milestones, Murderous Maths, Simon's Own Code, Space

Simulating Evolution: Evolutionary Steering Behaviors

 

On Wednesday Simon went on with playing god (evolution simulation) and translated Daniel Shiffman’s Evolutionary Steering Behaviors Coding Challenge from JavaScript to Java.  The goal of the challenge is to create a system where autonomous steering agents (smart rockets) evolve the behavior of eating food (green dots) and avoiding poison (red dots).

This challenge is part of the spring 2017 “Intelligence and Learning” course at NYU’s Tisch School of the Arts Interactive Telecommunications Program. Simon was especially happy to find out that Daniel Shiffman left a couple of personal comments praising Simon’s progress and offering help in pushing his code to Danniel’s GitHub repo.

Here is Simon’s translation on GitHub: https://github.com/simon-tiger/steering-behaviors-evolution

The rockets have their own DNA consisting of four genes:

IMG_4531

The challenge step by step:

 

 

 

 

 

Biology, Coding, Java, Murderous Maths

Autonomous Agents and Genetic Algorithms

Today Simon spent hours watching videos and reading Daniel Shiffman’s book The Nature of Code, concentrating on more complicated matters than ever: autonomous agents and evolution (genetic algorithms). In The Nature of Code there are two separate chapters covering these topics.  The term autonomous agent is defined as “an entity that makes its own choices about how to act in its environment without any influence from a leader or global plan”. This basically means that Simon has decided to try programming smart entities, that are like living things (have their DNA’s, behaviours, evolution).

The Autonomous Agent chapter also talks about steering force:

 

In the videos below, Simon has studied the steering behaviors in Daniel Shiffman’s code (Seek, Flee and Arrival) and changed the code slightly (third video) to be able to switch between the behaviours with a click of the mouse:

I also saw Simon go through the paper called Steering behaviours For Autonomous Characters by Craig W. Reynolds and reading about Persuit and Evade steering behaviours at this game development page.

In the video below Simon adjusted the Evolve Flow Field code to be able to see the possible behaviours (velocity vectors) of his smart rockets.

The code comes from the chapter Evolution of Code in Daniel Shiffman’s book, the chapter mainly devoted to genetic algorithms: it looks at the core principles behind Darwinian evolutionary theory and develops a set of algorithms inspired by these principles.

One of the most interesting notions Simon came across today was fitness, as in survival of the fittest. In the video you see Simon creating obstacles for the smart rockets. Together we observed how, as many generations of rockets passed, they learned to go around the obstacles better. This was possible because the rocket’s fitness was programmed to be greatly reduced every time it hit an obstacle.

All the rockets also have DNA’s: “We are marching through the array of PVectors and applying them one at a time as a force to the rocket”, Daniel Shiffman explains.

Simon also learned about genotypes and phenotypes, mating pool, crossover and mutation. He loved Daniel Shiffman’s example about haw many generations of strings with 18 random “genes” it would take to write “To be or not to be”.

Shakespear Evolution 18 Apr 2017