First Midterm Exam

The first midterm exam will be held in class on Friday, Oct. 17. The exam will be a written exam featuring short answer essay questions. The focus of the exam will be on the key concepts we have covered so far.

Topics to Review

Here is a list of the most important topics we have covered so far. You should be prepared to answer "what is this" or "how does this work" type questions about any of the topics listed below.

• Perceptron network
• Perceptron learning rule
• Linear layer in a pytorch network
• Activation functions
• Classifier output layer: probability distributions and softmax
• Loss function
• Key examples of loss functions: MSE, Cross entropy
• Data normalization
• Gradient descent/Backpropagation algorithm
• Convolutions
• Max pooling
• Dropout
• Residual connections

In addition to the topics listed above, you should also be prepared to discuss how a typical training loop works. Also be prepared to discuss the differences between training loops that use batches vs training loops that work with an entire training set at once.