in-class-exercise-assessments.md

MSc Coding 2 - In-class assignment submission.

Instructions - submit only 4 in-class exercises!

• Choose what you consider to be your 4 best in-class exercises that you have completed as part of Coding 2.
• Submit them in CODE format with an associated README file (one README per exercise), and any associated output image files. You can submit .cpp + .hpp files, iPython notebooks, any other code in text form, and example PNG/JPEGS (images in any format).
• For each of the 4 exercises, your readme should include a 250 word description of what you did, what you feel you learned, and most importantly, how the code should be run.
• Submit your 4 in-class assignments via MOODLE alongside the 4 README files in a zip file, with legible filenames.
• If you have git repos for your projects, please simply include a link to these in your zip file.

Here are details of the in-class exercises you were asked to complete. Of course, you should have spent some of your own time on these after each class, and so hopefully it won't take you too long to put the submission together.

Week 2 Exercise - Your first OF project

• Select a JavaScript project you completed last term and port it to C++ using openFrameworks

Week 3 Exercise - The Python Challenge!

• http://www.pythonchallenge.com

• Submit your python challenge solutions to the first 7 challenges

Week 4 Exercise - Python webscraper

• Build a simple webscraper that scrapes a set of documents from the internet and summarises them using gensim.
• If you manage to achieve this, extract keywords from all the different documents and see if any are more popular than others.
• Search for documents that contain those keywords using Python and then summarise those documents too.

Week 5 Exercise - Neural Networks by hand

• In week 5 we created a simple toy neuron by hand
• This should have left you with enough information to create a single later of neurons
• If you managed to do this, you may submit this as one of your in-class assignments.

Week 6 Exercise - Signal Processing in Numpy and Tensorflow

• NOTEBOOK: https://github.com/ual-cci/MSc-Coding-2/blob/master/Week-6-Exercise-intro-to-image-data-and-tensorflow.ipynb
• Make a version of the Notebook with at least one major difference that you have introduced yourself (as follows):
• First, you must do some transformation on the image dataset that isn't included in the above document. You must use numpy to do this transformation.
• If you manage to do this, your next task is to collect and process your own dataset instead of the one provided.

Week 7 Exercise - Using TensorFlow to create outputs

• Exercise 2
• Have a look at the following demos:
• https://research.google.com/seedbank/seed/neural_style_transfer_with_tfkeras
• https://research.google.com/seedbank/seed/deepdream
• Use your own images to create your own style transfer and deep dream outputs.
• Submit your image outputs. If you make any changes to the code, submit your own version of the notebooks, highlighting the changes you made using comments.
• You should also try to get the code running on your own machine. If you do this, submit the code that runs on your own device.

Week 8 Exercise

• Submit a port of any of the above exercises that will run on the Raspberry PI 4.