机器学习代写 | Machine Learning Assignment 6

本次美国代写主要为机器学习相关的assignment

Question 1: 10 points
In Keras download CIFAR10 dataset using keras.datasets.cifar10.load_data().
The dataset is composed of 60,000 3232-pixel color images (50,000 for training,
10,000 for testing) with 10 classes
Build a DNN with 20 hidden layers of 100 neurons each (that’s too many,
but it’s the point of this exercise). Use He initialization and the ELU activation
function. At the output layer use a softmax output layer with 10 neurons
Question 2: 10 points
Using Nadam optimization and early stopping, train the network.
Remember to search for the right learning rate each time you change the
model’s architecture or hyperparameters.
Question 3: 10 points
Add Batch Normalization and compare the learning curves: Is it converging
faster than before? Does it produce a better model? How does it a ect training
speed?
Question 4: 10 points
Try replacing Batch Normalization with SELU, and make the necessary adjuste-
ments to ensure the network self-normalizes (i.e., standardize the input features,
use LeCun normal initialization, make sure the DNN contains only a sequence
of dense layers, etc.).
Question 5: 10 points
Try regularizing the model with alpha dropout. Then, without retraining your
model, see if you can achieve better accuracy using MC Dropout.

Question 6: 10 points
Retrain your model using 1cycle scheduling and see if it improves training speed
and model accuracy.
Requirements
These are the general rules for submitting the homework not only for this as-
signment.
1. Submit jupyter notebook(s) in ipynb and html formats: html format allows
the grader to easily read it in canvas while ipynb allows to execute it if
something is not clear from html.
2. The notebook should be well formatted:
 Use markdown to break it into sections, bullet points, etc.
{ See l1_n3.ipynb for example how to use it.
{ If you double click on markdown cell, you can see the code.
{ To show the formatted text, execute the corresponding mark-
down cell.
{ Start the notebook with a markdown title cell of the form:
‘# Assignment X, Jane Doe, date’.
{ When answering a particular question from the assignment, start
the answer with markdown cell
‘## Question Y’
that indicates what question you are answering.
3. Name your notebook (and the corresponding html le) according to the
following template: Assignment_X_Jane_Doe.ipynb
4. If you are submitting several notebooks, append to the name ‘_partZ’
and explain in canvas what each notebook is about.
5. There should not be any failed cells in the notebook. Every cell should
work and have the results of the execution.
6. The notebook should contain only what is needed to answer a particular
question and nothing else. More is as bad as less.
7. All the plots, if needed for the assignment, should be of production quality
with readable labels, titles, etc. Also, try to nd the most approriate type
of plot to clearly demonstrate your statement.