Problem set 23
This problem set requires you to use this web site that
has a neural network
simulator. The options we will be fiddling with are the data
sets, the features, the number of hidden layers, and the number of hidden
nodes per layer.
- Use gaussian data, the first two features, 1 hidden layer with 1
neuron. Roughly how many epochs does it take to reach a test loss
of 0?
- Change to the xor data. After 2,000 epochs, what is the Test
loss? Is it correctly classifying most of the data? Does it look
like it is getting much better?
- What is the minimum number of hidden neurons required to quickly
(epochs < 1,000)
get nearly all the data right (test loss < 0.04) two thirds of the time?
- Switch to the circle data. Does your current network work well on
that data? How many epochs does it take to get a test loss less
than 0.01?
- Switch to the spriral data. Is your network working after 2,000
epochs? Try 8 hidden nodes. What is the test loss after 2,000
epochs. How many weights are in this network?
- Try 2 hidden layers of of 4 nodes each. What is the test loss
after 3,000 epochs? How many weights are in this network?
- Go full network, make both hidden layers have 8 nodes. What is the
test loss after 2,000 epochs? How many weights are in this network?
Try this 3 times and make a screenshot of the best one.
- Try going deep. Make a network with 6 hidden layers, with 3 nodes
each. What is the test loss after 2,000? How many weights are in
this network?
- Using the other features, design a network that gets a test loss
less that 0.1 within 2,000 epochs. Take a screen shot. How many
weights does it have?