May 18, 2022

This week we finally have started training our AI models! So it was more time-consuming than I thought, and my expectation has been not what I thought it would be. It’s actually more challenging to solve problems in machine learning because you can’t see the problem right away. Unlike in your typical software engineering where it gives you the errors right away so you can quickly work on it, in machine learning, you’ll have to train your models first before you can see how they perform. By the time you realize that there’s a problem, it can already be too late. So I was thinking that maybe one way to solve this is to prevent it from the beginning by writing a well-thought algorithm. Maybe that’s what math is for?

May 11, 2022

I just recovered from Covid and I didn’t have much time to work on our project. However, I’ve been doing research on ML Agents, like we’re supposed instead of adding more levels or features in our game, because we only have a few weeks left and Machine Learning might be too complicated to get started with with only a few weeks left, so that’s what I’m worried about. So I’ve been learning ML Agents through a Udemy course, but sometimes I doubt what I’m learning because we’re still learning about Genetic Algorithms, DNA Data Structures, etc that might not be applicable to our project. Although recently I’ve been learning about Perceptrons and I think that has something to do with Neural Networks, and I heard that Neural Networks is the algorithm that we might need specifically for our project. But we will see as I progress in my course!

May 3, 2022

I got Covid and tested positive last weekend so there’s nothing much to write because I couldn’t do much work. But I tried to write the Step-by-Step Instructions Document that were needed for our group assignment. Other than that, I had to ask our professor to give me an extension for the other work, especially for the Demonstration video since I can’t even speak properly and loud enough during recording because of my sore throat. I’m definitely getting better, but I’m also one week behind. I’ll catch up as soon as I get better.

April 26, 2022

After combining our prototypes, this week my team mates and I further developed our project. I did a thorough study of the code base and tested our game by playing it. While playing, I noticed that there were a lot of bugs so I tried to fix them one by one. All at the same time while my team is working on developing the other parts of the project, so I was also doing a lot of code reviews at the same time. Then, this week we also started setting up our environment for Machine Learning Agents in Unity. Since I have the new M1 Macbook, I had a lot of trouble installing softwares. Some softwares require certain versions, certain virtual environments, and much more pre-installations. There weren’t enough tutorials out there for M1 because it just came out, but there was one helpful website but it was in Japanese! So I had to translate it first, but at least it worked! At least, for now.

April 21, 2022

This week my team mates and I combined the prototypes we did individually. We did it individually for the purpose of learning the basics of C# and Unity. We decided to combine some parts like whose ball physics or paddle physics is better and integrate it into the base code. So first, we chose someone’s code as the base code and we just built it up from there. We decided to go with Ken’s code as the base code because we like most of his features, but we also added a lot of features from Jonathan’s code. We didn’t add much from mine but I’m thinking of adding some of the features from mine like when the bricks of different colors have to be hit for certain number of times and that it changes color when it hits, to the next more difficult levels in our game project. Overall I think this week has been the most complicated because of the integration and decisions.

April 13, 2022

This week my team mates and I worked on our initial project plan. It was tough creating ideas since everything is fresh from the start. In some parts, we had to put in something that we’re not even sure if we’re going to stick to or implement. Giving an estimate of the hours was really tricky as well since we don’t have any experience, we don’t actually know how long a feature could take to implement. Anyway, I also started setting up my environment for Unity but it was frustrating because there were a lot of dependencies that need to be installed, like the .NET package, figuring out the PATH for OmniSharp, and so on. But once I’ve got everything looking fine, I started playing with Unity with some basic sprites and positioning the visuals has been an ok experience so far. But for the physics, though, is actually kind of scary because my pieces just go everywhere! But I think this is where the fun will start though!

April 5, 2022

This week my team mates and I had a meeting and we worked on our team standards. We talked about future plans for our project: A Replica of Atari Breakout with Machine Learning!! I have no prior experiences with Unity nor C#, and neither anything about Machine Learning! So everything is new to me and I have to learn everything from scratch. One of my teammates has some experience so I hope he’ll be able to help us out about future problems. I wasn’t actually expecting to be put in this group because my first project choice was about the Language Learning App. I even intentionally took a Mobile Programming class prior to this class so that I could prepare well to make a Language Learning App for this class, but I didn’t even get to work on it. Anyway, the project we’re on right now seems like it would be fun though. Who wouldn’t be interested in making a game and apply A.I. to it? I guess I’ll just try my best, learn as much as I can, and most importantly, have fun with it!