Bill Xu reports on his experience at the Harvard University, Pre-Summer Camp, where he has been studying “Mathematical Elements of AI.” Bill says that his first week has been “both engaging and intensive.”
“We started by diving into linear algebra, a foundational subject for understanding and working with AI. We covered important topics such as matrices, vectors, and eigenvalues, and explored how these concepts are applied in machine learning algorithms. This solid grounding in linear algebra is essential for grasping more advanced AI topics,” Bill explains.
“In addition to linear algebra, we also began to understand backpropagation, the algorithm that enables neural networks to learn from their errors. By adjusting the weights to minimize the error in predictions, backpropagation enhances the accuracy of the model, which is crucial for developing more sophisticated AI systems,” Bill adds.
“Our learning was complemented by practical coding exercises, where we built simple neural networks to classify basic datasets. These hands-on activities reinforced the theoretical concepts we covered in class and provided valuable experience in applying these principles,” Bill says.
When not in class, Bill had a chance to make friends with several students, participate in fun activities on campus, and explore the city of Boston.
“Living in the dorms has been a pleasant and enriching experience. I share a suite with five other students, and we have quickly become friends. Our common room has been an excellent space for both studying and relaxing. We’ve spent a lot of time playing foosball and even sliding around the dorm room, which has been a fun way to unwind after classes. One of the highlights was a movie night where we watched “The Karate Kid” and enjoyed some delicious pizza,” Bill says.
“The weekend was filled with exciting social events. We visited an amusement park, where I had a great time riding the waterslides and competing in bumper cars. Additionally, we went on a tour of Boston, exploring historical sites and learning about the city’s rich history. These activities provided a perfect balance to the intensive academic schedule and allowed us to bond as a group,” Bill adds.
Week 2 — Core Principles
During his second and final week, Bill says the class focused on the “core principles of neural networks, which are essential for understanding modern artificial intelligence.”
“One of the critical concepts we covered is backpropagation, which is the algorithm used to train neural networks. Backpropagation is vital because it allows the network to learn from its errors by adjusting the weights in the network to minimize the difference between the predicted and actual outcomes. This process is essential for improving the accuracy and efficiency of AI systems. We also explored the workings of more advanced layers such as batch normalization and dropout,” Bill explains.
“Our final project is a group effort with five members, and it consists of three main parts. The first part involves working with the standard MNIST dataset, which contains handwritten digits. This section helps us grasp the fundamental concepts of image classification. The second part requires us to work with the CIFAR-10 dataset, which is more complex and involves classifying images of various objects, providing a greater challenge and allowing us to apply more advanced techniques. The third part of the project focuses on the Fashion MNIST dataset, where we will evaluate and fine-tune our model to achieve the best possible accuracy. This involves experimenting with different architectures, hyperparameters, and training strategies to optimize performance. Finally, we will present our findings and results in a final presentation. Working on these three parts provides a comprehensive understanding of neural networks and their applications in real-world scenarios,” Bill reports.
“Living in the dorms has continued to be a great experience. This week, we had a memorable movie night where we watched “The Karate Kid” and enjoyed pizza. We’ve also spent a lot of time playing foosball and sliding around the dorm room,” Bill adds.
“Overall, my time at the Harvard Pre-Summer Camp has been both productive and enjoyable. The rigorous academic curriculum, combined with engaging social activities, created a well-rounded and memorable experience. I gained valuable knowledge and skills in AI, formed lasting friendships, and enjoyed numerous fun and educational activities. This experience has not only deepened my understanding of artificial intelligence but also inspired me to pursue further studies and opportunities in this exciting field,” Bill says.
“I am grateful to the Garwin Family Foundation for supporting me throughout this journey. Their generosity made this incredible experience possible, and I cannot thank them enough for their unwavering support,” Bill adds.
We are so happy to hear that you learned valuable knowledge of AI, and were inspired to continue your studies in this cutting-edge field. Great work, Bill!
>> Read Bill Xu’s final report (PDF File, 70 KB)
>> Learn about the other students’ experiences in the GFF Scholarship Program.
