• University of Nottingham

Experimental Design for Machine Learning

Explore experimental design for machine learning in plant phenotyping, enhancing data collection and analysis.

African American Man Writing Lines Of Code On Desktop Computer With Multiple Monitors and Laptop in Creative Office. Male Data Scientist Working on Innovative Online Service For Start-up Company

Experimental Design for Machine Learning

  • 5 weeks

  • 1 hour per week

  • Digital certificate when eligible

  • Advanced level

Find out more about how to join this course

  • Duration

    5 weeks
  • Weekly study

    1 hour
  • 100% online

    How it works
  • Unlimited subscription

    $244.99 for a whole yearLearn more

Develop innovative strategies for machine learning with experimental design

Embark on a focused exploration into the core methodologies of experimental design tailored for machine learning, particularly within the context of plant phenotyping.

This course is designed to bridge the gap between theoretical knowledge and practical application, providing a comprehensive understanding of how experimental design principles can optimise machine learning outcomes.

Apply experimental design techniques in machine learning

Explore the foundational aspects of experimental design as it applies to machine learning. Understand the critical components of setting up experiments, from hypothesis formation to variable control and data analysis, which are crucial for achieving reliable results.

Investigate various experimental designs used in real-world machine learning scenarios, focusing on their applications in improving model reliability and performance.

Develop strategies for data collection and analysis for plant phenotyping

Delve into the strategies for effective data collection and annotation essential for training robust machine learning models.

Learn how to expand and refine datasets to cover a broad range of variables and conditions that will enhance the predictive power of your models.

Utilise techniques for model selection and performance

Sift through and select appropriate machine learning models and adjust parameters to maximise performance.

Discuss case studies demonstrating the successful application of these techniques in plant phenotyping.

By the end of this course, you’ll have a deep understanding of how experimental design supports machine learning, driving innovation in biosciences.

Syllabus

  • Week 1

    Experimental design

    • A photgraph of the lead educators on the course, Andy French and Mike Pound, sat together at a table

      Welcome to the course

      An overview of what's in store over the five weeks of the course

    • A stack of blurry photographs

      Data collection and annotation

      What are the things we need to look out for when collecting image data, and how can we add information to our images to help train machine learning algorithms?

    • An image of a leaf showing several different versions of a boundary drawn around itrid, with just two specific images highlighted as being of interest

      Summary and review

      Reflecting on what you have learned in Week 1, including a short quiz.

  • Week 2

    Understanding and working with data

    • A list showing examples of good filenames, and bad filenames

      Organising your datasets

      What do we need to consider in order to keep datasets organised? And how do we split data within datasets?

    • A screenshot of a 3D model of a field of wheat

      Expanding your dataset

      Tips on how you can expand the number of images in your training dataset, via augmentation, use of synthetic data, or other pre-existing datasets

    • An image of a plant, with it's own mirror image alongside itof the course

      Releasing data

      Why might you want to consider releasing your data?

    • A set of eight labelled plots arranged in a grid, each subdivided into three coloured sections

      Summary and review

      A review of week 1 of the course, plus a quiz

  • Week 3

    Choosing and using models

    • A still of Andy French and Mike Pound from one of the videos of the course

      Software, model selection, and training

      What to consider when choosing computing platforms and models for your machine learning projects, plus a look at training and inference.

    • Two diagram showing the same valley shape side by side. One is labelled 'small steps' and has an additional line snaking towards the bottom of the valley. The other had a line bouncing back and forth across the valley, never getting near the bottom.

      Improving performance

      What can we do to improve model performance?

    • An image of a wheat head, with a semi-opaque vector mask outlining the wheat head. This image is labelled "original". Alongside is the same vector mask, but with the original image background flipped horizontally

      Summary and Review

      A review of Week 3: Choosing and using models, including a quiz

  • Week 4

    Trusting results

    • A still of Andy French and Mike Pound from one of the videos of the course

      Trusting results

      How well can we trust the results of deep learning and machine learning models?

    • A confusion matrix for a classifier of four flower species, showing the numbers of true values against the predicted values for each of the three species

      Interpreting output

      How should we interpret the ouput from deep learning and machine learning?

    • A still of Mike Pound from one of the videos of the course

      Improving results

      Can we improve the results of our models?

    • rk symbol, over a grayscale background image of some plants

      Summary and review

      A summary and review of Week 4 - Trusting results, with a quiz

  • Week 5

    Practical tips and tricks

    • A large computing rack containing numerous GPUs

      Tips and tricks

      What to consider in terms of computer software and hardware, some tips on good practice when writing papers containing applications of machine learning and deep learning, and some thoughts on interdiscipinary communication.

    • A screenshot of a Colab notebook containing some computer code, an image of a flower with the output of an image classifier model shown below

      Course summary

      A wrap up of the course, with links to a practical exercise

When would you like to start?

Start straight away and join a global classroom of learners. If the course hasn’t started yet you’ll see the future date listed below.

  • Available now

Learning on this course

On every step of the course you can meet other learners, share your ideas and join in with active discussions in the comments.

What will you achieve?

By the end of the course, you‘ll be able to...

  • Design experiments that will collect good quality data for use in machine learning and deep learning models
  • Improve and organise datasets collected experimentally for use in machine learning and deep learning models
  • Compare and select machine learning and deep learning models for use with your experimental data
  • Interpret and evaluate results from machine learning and deep learning models and discuss ethical considerations surrounding their use
  • Report effectively on findings from machine learning and deep learning models trained on experimental data

Who is the course for?

This course is designed for bioscience professionals, particularly those in plant phenotyping, looking to enhance their skills in experimental design and machine learning to improve data collection, analysis, and model implementation.

What software or tools do you need?

No specific software is required. One demonstration will use Python.

Who will you learn with?

Michael Pound

Michael Pound is a Research Fellow and Lecturer in the School of Computer Science, University of Nottingham, UK. My research interests focus on Deep learning applied to plant phenotyping problems.

Andrew French

Andrew French is a Professor of Computer Science at the University of Nottingham. His area of research is developing novel image analysis methods, specifically for biological images.

Nathan Mellor

Nathan Mellor is a Post-Doc at the University of Nottingham. His research background is mathematical models of plants and plant tissues, using a range of programming and image analysis methods.

Who developed the course?

University of Nottingham

The University of Nottingham

The University of Nottingham is a research-focused campus university, described as “the nearest thing Britain has to a truly global university” by The Times. As a member of both the Russell Group and Universitas 21 international network, our reach extends across our campuses in the UK, China and Malaysia, with a diverse student body drawn from over 150 countries.

Ways to learn

Buy this course

Subscribe & save

Limited access

Choose the best way to learn for you!

$79/one-off payment

$244.99 for a whole year

Automatically renews

Free

Fulfill your current learning needDevelop skills to further your careerSample the course materials
Access to this courseticktick

Access expires 18 Mar 2025

Access to 1,000+ coursescrosstickcross
Learn at your own paceticktickcross
Discuss your learning in commentstickticktick
Certificate when you're eligiblePrinted and digitalDigital onlycross
Continue & Upgrade

Cancel for free anytime

Ways to learn

Choose the best way to learn for you!

Subscribe & save

$244.99 for a whole year

Automatically renews

Develop skills to further your career

  • Access to this course
  • Access to 1,000+ courses
  • Learn at your own pace
  • Discuss your learning in comments
  • Digital certificate when you're eligible

Cancel for free anytime

Buy this course

$79/one-off payment

Fulfill your current learning need

  • Access to this course
  • Learn at your own pace
  • Discuss your learning in comments
  • Printed and digital certificate when you’re eligible

Limited access

Free

Sample the course materials

  • Access expires 18 Mar 2025

Find out more about certificates, Unlimited or buying a course (Upgrades)

Sale price available until 3 March 2025 at 23:59 (UTC). T&Cs apply.

Find out more about certificates, Unlimited or buying a course (Upgrades)

Sale price available until 3 March 2025 at 23:59 (UTC). T&Cs apply.

Learning on FutureLearn

Your learning, your rules

  • Courses are split into weeks, activities, and steps to help you keep track of your learning
  • Learn through a mix of bite-sized videos, long- and short-form articles, audio, and practical activities
  • Stay motivated by using the Progress page to keep track of your step completion and assessment scores

Join a global classroom

  • Experience the power of social learning, and get inspired by an international network of learners
  • Share ideas with your peers and course educators on every step of the course
  • Join the conversation by reading, @ing, liking, bookmarking, and replying to comments from others

Map your progress

  • As you work through the course, use notifications and the Progress page to guide your learning
  • Whenever you’re ready, mark each step as complete, you’re in control
  • Complete 90% of course steps and all of the assessments to earn your certificate

Want to know more about learning on FutureLearn? Using FutureLearn

Do you know someone who'd love this course? Tell them about it...