Embedded Files
Learning Module 1
Learning Module 1
Soldering Practice
Justin and Chris spent the module learning how to solder based off of the tutorials
Learning Module 2
Learning Module 2
3D Printer - Lightsaber
Justin and Liam spent their module learning how to design a print 3d models. They designed their own lightsaber model and printed it using various methods to make it come to life.
NOT HIS IMAGE, USED THIS GUIDE
Learning Module 3
Justin spent the learning module relearning the basics of arduino and creating a stoplight function on the Arduino. He enjoyed it as it reintroduced him into the basics and got him wanting to use arduino more.
RPReplay_Final1702432930.movLearning Module 5
Learning Module 5
For this learning module, Justin and Paul worked together to make a working air hockey table with cool features using the Arduino. Their goal was to use the Arduino in two different areas, the first being the scoreboard and the second being the LED strip lights. For physical materials, the two used an air mattress fan, wood, Plexiglas, and Arduino kits to make the air hockey table. The LEDs ended up having multiple different programmed patterns while the scoreboard had programmed lights you clicked every time someone scored. They 3D printed the puck and sticks but had lots of trouble building enough pressure to make the "ice". Paul solved this problem by putting another piece of wood under the ice and used a pipe connected to the fan to build enough pressure. The air hockey table ended up working really well and it ended up being a really fun and educational learning module.
Learning Module 6
Chris and Justin took the Connect 4 that Chris made in his previous learning module and tried to implement AI into it, creating the perfect connect 4 bot. They used TensorFlow, Python, and ChatGPT and YouTube for research. They weren't able to accomplish what they wanted, but they were able to create a training program, learned about how Machine Learning works, and be able to play against the AI, even if it wasn't very good.
Here is the training function. This basically played AI vs AI multiple times to learn the correct moves and train the bot to be perfect.
Learning Module 7
Justin and Chris were inspired from a game that they both played, Rainbow 6 Siege, to create a drone similar to the one in the game. They tried to make a wireless controller using transmitters, and if possible wanted to connect a camera which could be transmitted wirelessly. Due to unfortunate circumstances with the transmitters, they were only able to make it move autonomously or with a wired joystick.
Learning Module 8
Paulie and Justin’s goals for this learning module were to use AI and machine learning to predict the outcomes of sports games, specifically NFL games. They achieved a lot with this learning module, starting with creating an AI that used data about the MAGIC gamma telescope. They were able to test the accuracy and loss when using artificial intelligence for predictions, and were able to import lots of different types of learning into the code. They used the Near Neighbors strategy (kNN), Naive Bayes, Logistic Regression, an SVM, and neural networking. In the neural network, Paulie and Justin made a test with different learning rates, batch sizes, number of nodes, and dropout probabilities which method had the least validation loss. Because they were constantly absent with doctors appointments, ap tests, and peer events, Justin and Paulie fell short of their original goal. They didn’t really know where to start as neither of them had worked much with AI, but watched lots of different tutorials from different creators using different strategies to use AI. They started watching someone use Visual Studio but realized that seemed a little too ambitious, then went on to watch a girl who used Jupyter Notebooks, and finally found the perfect video where a lady used Google Collab to use AI on various datasets. It was hard adjusting to their own dataset afterwards as they had trouble finding their own dataset that could upload, but everything worked out in the end.
Here is the set of NFL data they used
Report abuse