COMP5329 – Deep Learning
Assignment-1
Due: Thursday, 13 April (Week 7)
1. Task description
Based on the codes given in Tutorial: Multilayer Neural Network, you are required to accomplish a multi-class
classification task on the provided dataset.
In this assignment, you are expected to implement the modules specified in the marking table.
You must guarantee that the submitted codes are self-complete, and the newly implemented modules can be
successfully run in common python environment.
You are NOT allowed to use Deep Learning frameworks (eg PyTorch, Tensorflow, Caffe, and KERAS), or
any kinds of auto-grad tools (eg autograd).
Scientific computing packages, such as NumPy and SciPy, are acceptable.
If you have any question about the assignment, please contact: Mr Jinxu Lin [email protected]
Dataset
The dataset can be downloaded from Canvas. There are 10 classes in this dataset. The dataset has been split
into training set and test set.
2. Instructions to hand in the assignment
2.1 Go to Canvas and upload the report. The report should include each member’s details (student ID and
name).
If you work as a group, only one student needs to submit the report which must be named as student ID numbers
of all group members separated by underscores . instructions on how to run your code in the appendix of the report or include a readme.txt in your shared folder. Don’t update the code/data any more after the submission. If the latest modified time of the shared folder is signiftanly late after the submisison deadline, the whole submission will be taken as a late submission. 2.3 The report must clearly show (i) details of your modules, (ii) the predicted results from your classifier on test examples, (iii) run-time, and (iv) hardware and software specifications of the computer that you used for performance evaluations. 2.4 There is no special format to follow for the report but please make it as clear as possible and similar to a research paper. 2.5 If you use ChatGPT or other AI tools for the assignments, please clarify how you have used them in the report. Late submission: Suppose you hand in work after the deadline: If you have not been granted special consideration or arrangements
– A penalty of 5% of the maximum marks will be taken per day (or part) late. After ten days, you will be
awarded a mark of zero.
– eg If an assignment is worth 40% of the final mark and you are one hour late submission, then the maximum
marks possible would be 38%.
– eg If an assignment is worth 40% of the final mark and you are 28 hours late submitting, then the maximum
marks possible marks would be 36%.
– Warning: get submission sites very slow near deadlines
– Submit early; you can resubmit if there is time before the deadline.
3. Marking scheme
Category Criterion
Report [50]
Introduction [5]
– What’s the aim of the study?
– Why is the study important?
Methods [15]
– Pre-processing (if any)
– The principle of different modules
– What is the design of your best model?
Experiments and results (with Figures or Tables) [20]
– Performance in terms of different evaluation metrics.
– Extensive analysis, including hyperparameter analysis, ablation studies and
comparison methods.
– Jusitification on your best model.
Discussion and conclusion [5]
– Meaningful conclusion and reflection
Other [5]
– At the discretion of the marker: for impressing the marker, excelling
expectation, etc. Examples include fast code, using LATEX, etc.
Modules [45]
More than one hidden layer [5]
ReLU activation [5]
Weight decay [5]
Momentum in SGD [5]
Dropout [5]
Softmax and cross-entropy loss [5]
Mini-batch training [5]
Batch Normalization [5]
Other advanced operations (eg, GELU, Adam) [5]
* Please make a highlight if you have one you think are advanced.
Code [5] Code runs within a feasible time [5]
Code Penalties [-]
Well organized, commented and documented [5]
Badly written code: [-20]
Not including instructions on how to run your code: [-30]
Late submission