Skip to main content
Department of Information Technology

Deep Learning - Hand-in Assignments

During the course 4 hand-in assignment (HA) will be completed.

Hand-in assignment 1A

Implementation of a logistic regression model and training with gradient descent. Implementation in numpy/Python or similar language of your choice.

Hand-in assignment 1B

Implementation of a forward neural network and training with mini-batch gradient descent. Implementation in numpy/Python or similar language of your choice.

Hand-in assignment 2

Convolution neural networks and regularization. Classification and segmentation. Implementation in Tensorflow/PyTorch/Keras or similar.

Hand-in assignment 3

Deep time series models. Implementation in Tensorflow/PyTorch/Keras or similar.

Programming language

To solve the assignments in the course you may use any software you want, but we recommend to use Python. You do not need any knowledge about Python to complete the assignments in the course, but you do need some programming experience from other languages for scientific computing like Matlab or R and be willing to learn it on your own. We recommend to install Python 3.6 via Anaconda. Then you get both Python and Jupyter Notebook which you need to run the notebook, as well as other commonly used packages for scientific computing and data science. If you are new to python the following introductory crash course is also very well suited for the content of this course.

Regulations for the Hand-in assignments

  • Each HA will come with a deadline. Obviously you are supposed to submit until the deadline. However, there are some exceptions to this, explained below.
  • Everyone has a total of 7 joker days which can be used with any HA. Using a joker day with a HA means that you can submit the corresponding HA 24 hours later than the deadline. Joker can only be used in unit of days. A joker day affects the deadline of the corresponding HA only.
  • In the report, you are supposed to give your plots, results, comments, interpretations, speculations, problems encountered during the implementation and how they are solved etc. We also want your commented code as a part of your report in an appendix. The report should be a pdf document.
  • For all employees at Uppsala University, submissions will be made via the student portal. After logging in to the student portal with your employee account you find the course under the tab "my internal trainings" or by following the link above. If you are not employed at Uppsala University (MSc students and external PhD students), submit your report per email to Niklas Wahlström with the pdf file of your report for the HA. Name the file [first name]-[surname]-HA[x].pdf, for example Niklas-Wahlstrom-HA1A.pdf.
  • You are encouraged to collaborate with other course participants but you should write and submit your own report and code. If you did collaborate with others, state with whom in report.
Updated  2020-12-15 10:34:40 by Niklas Wahlström.