<p>I'm about 2/3rds of the way through my ISEF/Siemens project. I have about 1/2 the paper written(the other stuff will be data analysis, conclusion, possible extensions, etc), and about 3/4 of my actual code base completed.</p>
<p>I was hoping someone could take a look at what I have of my paper, and give me any advice. At this point in the project, it is possible to change stuff, but once I close out the code and start producing data, it will be difficult, since the results is dependent on time. The more time I give it, the more relative the results will be. Furthermore, I have no idea how much time the species has to evolve for, so its kinda tricky. </p>
<p>Please? The topic is AI. Here is the abstract:</p>
<p>This experiment is conducted via a simulation. The simulation consists of simple environment, and a virtual species. Each member of this virtual species has the same physical attributes. Members of this species have the ability to hurt each other(aim and fire a gun), so that in order to survive, the bots must form intelligent plans and possibly work together.</p>
<p>Yes, very cheesy, but I think that if you read how I aim to achieve this, it might be not as cheesy.</p>
<p>Give it a real corker for a name, something like, "A hypothesis to understand the semantics of species behavioural patterns when confronted with a hostile environment and provided with self defence heuristics"</p>
<p>Damn Havaldaar. That seems a thousand times better than my lame abstract. I read that once, and it gives me the impression that I get when people are like, calculating the accuracy of spectrum analysis on diverging galaxies. Thats pretty awesome. </p>
<p>Do you know where I can learn how to write like that? Should I just read scientific papers?</p>
<p>Learn to write like that!LOL, it was a joke, I love making up nonsense technical names. Did you read the new item on how some tech college kids sent in some scientific sounding rubbish to a Science conference as a prank, and they hosts actually included it in the program of events!</p>
<p>sagar, I cant really help you on your project but maybe you can help me? I am interested in the same field as you but Im not really sure what to do for my ISEF project (i still have 2 years to go) can you maybe email me with info on your project? <a href="mailto:argunda@gmail.com">argunda@gmail.com</a></p>
<p>Sure Gandhiji. Would it be ok if I sent you a copy of the paper in a week or two. Right now I'm experimenting with it in Word, and kinda messed up the formatting.</p>
<p>Havaldaar: That paper generator(scigen i believe) was developed by MIT grad students. I read that it is based on Mad libs, as they wrote a few hundred sentences, and interchanged nouns, verbs, adjectives, etc.</p>
<p>Another question that has occured to me: What do they consider in the Intel project? If my project is a good idea, but I don't get successful results, am I thrown out? </p>
<p>My experiment is kind of like the progression of a species. Obviously if you took a look at mankind at 10 BC and compared it to 2005 AD, there is a difference. Similarily, my experiment works on the principle that the more time spent computing, the more likely good results will be seen. Furthermore, the data has to be very closely analyzed, ie you have to sit and watch the bots living their lives like a sitcom. So the results are subjective.</p>
<p>Also, for Siemens Westinghouse, my partner sucks, and my plan is to just plug him along until the point we both get interviewed, and then I'm screwed.</p>
<p>I'm using C++ for the main code. The computations are so expensive, real time viewing is impossible. So rather, I record it to disk(just what I need), and then play it back like a movie.</p>
<p>Haven't decided how to write the movie editor yet.</p>
<p>NEVER i repeat NEVER have a partner for a science fair. My partner screwed our projuect up the last time around( although in all fairness it was pretty bad to begin with.)</p>
<p>And about your question, it doesnt really matter if your results are not succesful. Disproving a hypothesis is just as important as proving it, and has often led to greater scientific discoveries. However you should show everything in your abstract and make it clear that your results didn't support your hypothesis.</p>
<p>No no no. Not like disprove or prove your hypothesis. I mean that the experiment never ends. Even if you saw no signs of intelligence, you could have run it for a few more hours, and things could have changed.</p>
<p>Well i suppose you'd have to extrapolate(another humbug of a scientific word, take notes down lad and youll go far ;)) the observations. Conjecturing is a huge part of Sience...my whole Chem 12 th grade portion seems like conjecture.</p>
<p>LOL I remember in the 11th we did Bohr's Atomic theory and we realized that the disadvantages to the theory were far more than the advntages and my friend shouted out,"Bohr was such a shameless scientist!" The Chem teacher gave us looks that could kill.</p>
<p>reading many scientific papers really improves your writing and jargon in your field of study. Try reading some computer science papers from journals.</p>
<p>So for Siemens, I have to enter as a team. Lets say that I put in me and my friend. Now what do they look at? Do they contact us individually? At what point do we both get questioned? Because this guy doesn't know what my experiments basically. He knows the gist, but can't describe some of the details.</p>