07 Prove : Assignment
Neural Network Part 2
Objective
Understand Neural Networks, in particular, back-propagation for multi-layer perceptron networks.
Please note that you have 3 weeks to experiment with this algorithm. This is because it can be a challenge! This is likely the most challenging individual programming assignment we will do this semester. Do not put this off!
The purpose of this assignment is to make sure you are on track to be successful with this assignment.
Instructions
This week your task is to use a library implementation to make predictions on several different datasets of different characteristics. Then, next week, you will pick one or two of these datasets and experiment with various hyper-parameter combinations.
In particular, you should use and make predictions on at least four different datasets. Then, among your datasets you should make sure that you have worked with at least the following characteristics in some form:
Text values
A significant amount of columns
A significant amount of rows
Binary classification (only two possible values)
Classification with more than two target values
Regression (numeric target values)
Look at different measures of success for your datasets (more than just accuracy)
This does not mean that you must use a different dataset for each of the above characteristics, nor does it mean that one dataset needs to cover all of them. You may have one dataset that covers a few, and other that covers others.
In the end, your goal is to have prepared and made predictions on a number of different datasets so that you can choose a "hard" one or two from among them that you will focus on next week.
It is your task to find these datasets. One potential resource to consider is the UCI Machine Learning Repository from which many of our datasets this semester have come.
Submission
Because this is a milestone toward next week's assignment, this assignment does not specifically require (or reward) going above and beyond. That will be addressed in the assignment for next week.
You are, of course, encouraged to go beyond these deliverables because that will help you finish the assignment for next week.
When complete, answer the questions in the accompanying I-Learn quiz.