BUSOBA 7247 Spring 2023 AI and ML Problem Set 1
Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit
Problem Set 1
AI and ML
BUSOBA 7247
Spring 2023
Recommended HOML Questions
Note: You do not need to turn in your answers to the problems from
HOML. For solutions, see homl.info.
– Chapter 1: Exercises 1, 3, 4, 8, 12, 14, 16, and 19
– Chapter 3: Exercise 1
Assignment Questions
Please turn in an HTML export of the Jupyter notebook you use to
solve these questions. Include answers to the specific questions as
markdown cells or Python comments.
Building on the application of k-nearest neighbors to the MNIST
handwritten digits dataset (see 02_knn_mnist .ipynb):
1. Use automation to find a good value of k for multiclass classification via cross-validation.
2. How would you characterize the performance of the model you selected (i.e., the nearest-neighbors model with your value of k)?
3. Would you expect your model to generalize to new instances similar to those from the MNIST handwritten digits database?
4. Describe the objective of the narrow AI you’ve created.
5. Going beyond the MNIST example: consider Stuart Russell’s ar- guments regarding the definitions of objectives and intelligence (see, e.g., the passage under the heading“What Went Wrong?” on page 9 ofHuman Compatible). Discuss a situation wherein an AI could achieve its objective —i.e., the objective the AI’s creators specify —yet the AI’s actions do not align with the creators’or soci- ety’s goals.
2023-01-29