- Project 1: Propose a change in line 18 of [Program 2-3]
- Project 2: With the handwritten numeric dataset provided by sklearn, learn from the following two models and measure performance.
2-1 Modify [Program 3-5] 2-2 SVM: Adjust C and kernel hyperparameters 2-3 Random forest: adjust n_estimators and max_depth hyperparameters - Project 3: Build a convolutional neural network classifying CIFAR 100 using Tensorflow.
3-1 Experiment with multiple neural network structures and multiple hyperparameters to produce the highest performance model.
3-2 Select two to three types of hyperparameters that you think are important and optimize them.
3-3 Use 5-layer cross-validation for performance experiments. Measure top-1 and top-5 accuracy - Project 4:Create interesting applications with the face detection, face mesh detection, hand detection, and posture estimation capabilities provided by MediaPipe. Make UI using PyQt.