Online learning (5cr)

Evaluation scale

0 - 5

Objective

The student knows how to create a neural network for online learning, as well as enable it on the server and publish it. The student can adjust the hyperparameters of the neural network so that the neural network learns more from user data. (E.g. recommender neural networks or congestion predicting neural networks.)

Location and time

Blended learning.

Materials

The IaC Terraform workshop, which you will study during the first weeks of the course, is available online: https://gitlab.dclabra.fi/workshopit/iac-terraform

The access to the rest of the study material is only to those who partake the course. It is in GitLab and in YouTube. Also, the Terraform and Azure documentation will serve you during this course.

Teaching methods

Pre-recorded lessons or tutorials, exercises and Teams Classroom support sessions.

In this course, you will get to know the Terraform tool and the Azure cloud service (e.g., Azure Storage, Azure Container Instances, Azure Container Registry) as well as Azurite. You will have access to a corresponding service created by the instructor. You will receive both the code and videos in which the service is assembled. Additional support is provided in Teams Classroom sessions.

Assessment criteria, satisfactory (1)

For a grade of 1, it is required that the returned course exercises show that the student knows how to use a ready-made neural network in an online environment.

Assessment criteria, excellent (5)

A grade of 5 requires completing and returning all course exercises, as well as a commendable reflection on the exercises.

Qualifications

Deep learning 1
Deep learning 2

Further information

In the assignments, you may use AI for exploring solutions and supporting your learning. Any code written by AI must be tested, commented, and its working principles understood. Keep in mind that you must be able to explain your solutions in the recorded demo. Please do not outsource your brain to AI.