Coursework 1 description
The aim of this coursework is to explore the RMSProp and Adam learning algorithms (discussed in lecture 5) in the context of L2 regularization (discussed in lecture 4). The motivation for the coursework comes from the recent paper by Loshchilov and Hutter on using the Adam optimizer with L2 regularization and weight decay. The coursework uses an extended version of the MNIST dataset, the EMNIST Balanced Dataset.
The code for the coursework is available on the course Github repository on a branch
mlp2018-9/coursework_1. This also includes a LaTeX template (see the
report directory) for the report that you will need to submit. Full details in the Coursework 1 document.
Coursework 2 description
The aim of this coursework is to further explore the classification of images of handwritten digits using neural networks. As in the previous coursework, we'll be using an extended version of the MNIST database, the EMNIST Balanced dataset, described in the Coursework 1 description. The first part of the coursework will concern the implementation and experimentation of convolutional networks using the MLP framework. The second part will involve exploring different convolutional network architectures using PyTorch.
In order to support the experiments you will need to run for the second part of the coursework (which will be carried out in PyTorch we have acquired Google Cloud Platform credits which allow the use of the Google Compute Engine infrastructure. Each student enrolled on the MLP course will receive a $50 Google Cloud credit coupon which is enough to carry out the experiments required for this coursework. You will receive an email which will give you the URL you will need to access in order to request your Google Cloud Platform coupon.
The provided code and setup information for this coursework is available on the course
Github repository on a branch
mlp2018-9/coursework_2. This also includes a LaTeX template (see the
report directory) for the report that you will need to submit. Full details in the Coursework 2 document.
MLP in semester 2 will be based on projects done in groups of 2-3 students. The projects, which should be done using one of the deep learning toolkits (TensorFlow or PyTorch are recommended) can be chosen from a variety of topics and data sets. These projects correspond to courseworks 3 and 4.
Please remember the University requirement as regards all assessed work for credit. Details about this can be found on the page describing the academic misconduct policy.
Furthermore, you are required to take reasonable measures to protect your assessed work from unauthorised access. For example, if you put any such work on a public repository then you must set access permissions appropriately (generally permitting access only to yourself, or your group in the case of group practicals).
Copyright (c) University of Edinburgh 2015-2018
The MLP course material is licensed under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International Public License
The software at https://github.com/CSTR-Edinburgh/mlpractical is licensed under the Modified BSD License.
This page maintained by Steve Renals.
Last updated: 2019/09/16 15:42:09UTC
Informatics Forum, 10 Crichton Street, Edinburgh, EH8 9AB, Scotland, UK
Tel: +44 131 651 5661, Fax: +44 131 651 1426, E-mail: email@example.com
Please contact our webadmin with any comments or corrections. Logging and Cookies
Unless explicitly stated otherwise, all material is copyright © The University of Edinburgh