Curriculum overview

See the curriculum details here.

Structure

The Starters Academy is made of one Bootcamp, 6 specialisations, and 1 capstone.

Bootcamp

The bootcamp is the launching event of the Academy, where students attend online classes with Academy instructors.

Specialisation

A specialization is a set of learning units and a hackathon, centred on a particular topic (e.g. time series). The first specialization consists of 14 mandatory and 2 optional small learning units. The other specializations have 3 big learning units. All specializations end with a hackathon.

Hackathon

A hackathon is a 1 day event where students, in teams, solve a data science problem. The students receive a dataset and have to set up and optimize a model. The teams competes for the best model performance as measured by a single score. Teams are assigned at random, as students also need to learn how to work with different types of people.

Capstone

The capstone is the final project of the Starters Academy. It requires each student to individually execute a project, which will be graded. See more details under Capstone Project.

Learning units

A Learning Unit is composed of the following:

1. Learning Notebooks (Jupyter Notebooks)

The high granularity teaching should happen in the Learning Notebooks. These should be sufficient to teach the materials on their own and serve as a reference. The Learning Notebooks teach only new concepts, and refer to previously learned concepts to the respective learning units. A final section called “If you want to learn more (optional)” is included in some notebooks to guide interested students in further studies.

2. Exercises with expected output (Jupyter Notebooks)

Exercises are notebooks that cover the material taught in the same learning unit’s Learning Notebooks. They are based on the Coursera idea of providing an expected output and ensuring the student only needs to worry about a highly granular piece of material.

Most exercises will demand an exact answer while exercises which more concerned with workflow may have a slightly more open objective, such as exceeding a certain performance metric. Exercises should be done individually, and are graded by an auto-grader.

Example of exercise with expected output:

Any learning unit requires a grade of at least 80% to be considered passed, with an unlimited number of attempts (but with a deadline).

3. Presentation

Presentations only happen in Specialisation 1 in the form of virtual classes. Presentations are short, motivating introductions to the topic, which should explain only the concepts and provide insights without going into implementation details.

The main objective of a presentation is to ensure that the students understand “why” they are learning a topic, rather than teaching the full topic.

Big vs small learning units

Learning units are modular and serve to teach highly related concepts. There are two types of Learning Units:

1. Small Learning Units (SLU)

SLUs are taught in Specialisation 1 and are accompanied by a presentation. The expected time split for students is 1 - 2 hours per unit.

2. Big Learning Units (BLU)

BLUs form the backbone of the Academy after the Bootcamp ends. They do not contain presentations, as they are done in remote mode. The total time a student is expected to spend on a BLU is 5 - 10 hours.