Practical Implementation and Written Report for Machine Learning Project

Requirements for Practical Implementation

1. Practical Implementation (complete code, trained model and data to replicate the results). The source code must be error-free (i.e. no debugging necessary to run). Minimum requirements:

• Coherent implementation (no errors, sound training pipeline),

• Appropriate metric (Do not confuse metrics for classification,regression, clustering, dimensionality reduction, etc.)

• Sound evaluation (comparison with trivial baselines, such as random flipping coin, outputting the most frequent label, outputting the previous value in time series, and existing machine learning methods that can be applied to the topic.).

2. Written Report (4 pages of a detailed description of the steps taken in carrying out the project):

• Must follow the template uploaded on the Moodle website.

• Should include references (citing other work) where appropriate (when images, data, code, or any other resources have been used from other sources)

• For document structure, please follow the template.

The coursework marking will be based on the report, in principle. The submitted source codes are used, for example, to verify whether the description in the report reflects the implementation and the appropriateness of the implementation but do not have a specific ratio in marking. Even if the implementation is of good quality, it will be ignored if it is not explained well in the report.