Skip to main content

Introduction

Introduction
Complete

What is the best way to build a machine learning model?

Is it to employ neural networks to leverage connections in the data?

Or is it to apply complex transformations to ensure the data is in the most appropriate format before building?

Depending on your data and what you want to achieve, the method you use to build your models will vary. This is why it is important to understand the various model architectures and training methodologies, so we can select and implement the most appropriate solutions for our organisations.

In this unit, we will explore these concepts and understand when they are best used.

Hand pointing illustration

Why does this unit matter?​

Machine learning (ML) is not a one-size-fits-all approach. Your data, organisational infrastructure and business goals are all key components in the process.

It is important that you understand what different methodologies are available and be able to critically evaluate which is the most suitable for your goals.

This will allow you to more effectively justify your decisions and explain insights to key stakeholders and decision-makers.

Learning objectives​

By the end of this unit, you will be able to:

  • Analyse various ML model architectures and their suitability for different types of business problems, considering quality, efficiency, and validity principles.
  • Evaluate advanced training methodologies and their impact on model performance, security, and generalisation across diverse datasets.
  • Develop frameworks for selecting and deploying appropriate ML models and hardware architectures based on computational requirements and available resources.

Before you continue, make sure you've completed the following units:​

  • Module 6 Unit 1: Model Engineering and Training Fundamentals
  • Module 6 Unit 2: Training Process Optimisation

Action item: Pause and think​

As we get started, reflect on the following questions to connect your prior understanding to the applications this content will introduce.

3 / 3 questions
What do you consider to be best practices for building machine learning models?

understand the data first , clean up the data , prepare the data and understand what should be our outcome

Are you aware of any hardware limitations in your organisation that might limit your ability to deploy the most powerful models?

compute is finite and while we can use cloud , its not sure what hardware we use to train the models if any

What do you think the process should be to select and deploy the most appropriate model for your goals?

We need to have a balance between compute , speed and output .