COS7054-B - Cloud Artificial Intelligence Solution for a Real-life Problem Individual Research Project: Cloud AI, Coursework 2

Aims

1. Demonstrate appropriate use of applicable Cloud data analytics and programming frameworks to implement AI applications.
2. To implement a prototype AI system that can investigate the dataset used for your case study.
3. Create and save an ML model Pipeline (Sklearn, TensorFlow, etc.) as a ‘production ready’ function.
4. Critically analyse and select appropriate techniques and algorithms depending on the context and problem.
5. Discuss the legal and ethical aspects of your research problem.

Problem statement

As a member of the Big Data team, you aim to demonstrate the capability of AI and cloud environments in solving real-life problems. This should be done by working on an individual case study in which you will show how AI and cloud frameworks can solve or improve it.

•  Explore it using well-established, high-quality references and statistical functions.
•  Implement an ML pipeline using any ML framework.

For this coursework, the requirement is to produce a document (4000 words) on a piece of research you have undertaken to describe the cloud AI framework and method for the case study.

The aim is to investigate your research problem showing the role AI and cloud services can play in automating and optimising processes, analysis, and detection. In doing so, you are expected to perform the following tasks:

• Select a relevant benchmarked data resource(s).
•  Investigate your dataset and have a basic understanding of its significance and meaning.
• Explore it using well-established, high-quality references and statistical functions.
•  Implement an ML pipeline using any ML framework and save it.
•  Propose a cloud AI system or framework that can be implemented for your chosen dataset.
• Describe and critically assess AI services (including security aspects_ on the cloud, showing the spectrum of Cloud computing capabilities used to deploy them.
•  Discuss your findings, making recommendations for further investigation and improved implementation.
• Write a 4000 words report to present your work and findings.
• Wire a piece of code showing your AI model and your saved model/models.