Submit your Jupyter Notebook for this class’s final project. Follow the directions below.
1. Explain your process. What will your machine-learning pipeline look like? Explain the process that you
will follow. How will you train and test your model? When doing this, also explain and justify your scoring
tecnique. For example, for home value, you may not actually want to use a symmetric loss function. For
predicting appeal, you may not want to use classification accuracy.
2. Even though you are interested in prediction and not inference, you still need to decide what data you
will use. If you use only data already in this dataset, explain why you chose the features you chose. If you
add new features, explain why you are incorporating them and how you got them.
3. Create the best models that you can. Explain how you choose its best parameters and how you
evaluated it against competing models.
4. Explain the implications of your models and what steps you could take next to improve them.
You will be graded on:
Data preparation. Did you appropriately prepare your data for modeling?
Your process and pipeline. Are you taking appropriate steps to train and test your models?
Model creation and selection. Are you creating, tuning, and selecting models aprpopriately?
Your explanations. Did you appropriately explain your rationale for the decisions you made and for
the conclusions you reached?
Purchase answer to see full