In from three to eight years we will have a machine with the general intelligence of an average human being. — M. Minsky, Life Magazine, 1970

You will note that Google provides solutions for some of these exercises; be sure that you understand them and that you answer the extra questions included in this lab assignment.

Pandas

Finish the Pandas tutorial specified in the guide.

Exercise 7.1

Intro to Pandas — Complete these exercises (cf. the guide).

Questions:

  1. Submit your solutions to exercises 1–2.
  2. Why would one use Pandas rather than the standard data manipulation features provided by NumPy?
  3. Under what circumstances would it be useful to reorder/shuffle a Pandas DataFrame?

Save your answers in lab07_1.txt.

TensorFlow

Finish the First-Steps-with-TF programming exercises. Note that we will primarily be programming neural networks using Keras, so you don’t need to fully understand the TensorFlow code.

Exercise 7.2

First Steps with Tensor Flow.

Questions:

  1. Compare and contrast categorical vs numerical data
  2. Submit solutions to tasks 1–2. Include your best hyper-parameter values and the resulting RMSE, but not the training output.
  3. What are the hyper-parameters learned in these exercises and is there a “standard” tuning algorithm for them? Explain your answer.

Save your answers in lab07_2.txt.

Here, we experiment with synthetic features.

Exercise 7.3

Synthetic Features & Outliers

Questions:

  1. Submit solutions to tasks 1–3.
  2. What is the purpose of introducing synthetic features?
  3. What are outliers and what is typically done with them?

Save your answers in lab07_3.txt.

Checking in

We will grade your work according to the following criteria:

See the policies page for lab due-dates and times.