Ideas

The concepts, tools and methods that make data science possible. In this section: Tutorials | DataScienceBites

Our ideas content is in development. Interested in contributing to Real World Data Science? Check out our call for contributions.

An image of the cover of the RSS Data Science and Artificial Intelligence

RSS: Data Science and Artificial Intelligence - showcase your research
AI
Data Science
Machine learning
Deep learning
Econometrics

RSS: Data Science and Artificial Intelligence is now open to submissions. It offers an exciting venue for your work in these disciplines with a broad reach across…

Feb 5, 2025

Close-up of a copy of the Nobel Prize Medal. Photographed on the floor of the Nobel Museum in Old Town, Stockholm. Machine learning came up a winner in both the Physics and Chemistry Nobel Prizes for 2024. Credit Shutterstock.

The machine learning victories at the 2024 Nobel Prize Awards and how to explain them
AI
Algorithms
Machine Learning

Machine learning using apparently different architectures bagged the 2024 Nobel Prizes for Physics and Chemistry. Anna Demming reports on what the prizes were awarded for…

Anna Demming
Oct 31, 2024

Muted. In the absence of proactive intervention, the shift of debate and discourse online risks muting women and girls as multiple factors exclude them from engaging there as productively as male counterparts. Copyright Park Kang Hun/Shutterstock.

Are we at risk of muting the female voice in the digital world?
gender equality
skills
ethics

Anna Demming reports on recent insights into the interplay between gender disparities on and offline and what might help to close the gap.

Anna Demming
Sep 17, 2024

Growth of GDP with statistical graph, 3d rendering. Digital drawing. Credit: Shutterstock, Vink Fan

Nowcasting upgrade for better real time estimation of GDP and inflation
Forecasting
Machine learning

Atmajitsinh Gohil describes a recently devised 3 step framework that improves the accuracy of nowcasting.

Atmajitsinh Gohil
Jun 25, 2024

Themis, goddess of justice. External governance is required to ensure the outcomes of AI deployment are safe and just. Credit Shutterstock, Michal Bednarek

AI series: Ensuring new AI technologies help everyone thrive
AI
AI ethics
Regulation

This AI article series has highlighted a number of areas where AI can offer genuine benefits, as well as flagging how best to dodge the pitfalls and cultivate models and…

Anna Demming
Jun 11, 2024

The panel clockwise from top left: Ali Al-Sherbaz, Janet Bastiman, Fatemeh Torabi and Jonathan Gillard

AI series: What is “best practice” when working with AI in the real world?
AI
large language models
machine learning

Working with AI in real world conditions can be quite a different proposal to the idealised settings often discussed “in theory”. Guest editor Anna Demming speaks to a panel…

Anna Demming
Jun 4, 2024

The Industry of Ideas Infrastructure (provided by Jason Owen-Smith, IRIS, University of Michigan)

AI series: Meeting the unprecedented challenges AI poses in the labour market
AI
Data management
Forecasting

Expanding the deployment of Artificial Intelligence (AI) across various sectors has been widely touted as a significant threat to jobs but so far data to justify these fears…

Julia Lane, Lesley Hirsch, and Adam Leonard
May 28, 2024

Evaluation plots. The kinds of charts of monitored performance and risk metrics that are plotted to evaluate an AI model. Reliable deployment of an algorithm requires identifying appropriate metrics for performance and risk as well as rigorous, ongoing evaluation. Image created by Isabel Sassoon using Adobe Firefly to show a technical report process flow of statistical model performance and a huge numbers chart.

AI series: Evaluation essentials for safe and reliable AI model performance
AI
Machine Learning
Large Language Models

The real world can be a dangerous place for an AI algorithm, full of unforeseen data, evolving behaviours and significant human consequences. Isabel Sassoon highlights some…

Isabel Sassoon
May 21, 2024

Eleanor Roosevelt reads the Universal Declaration of Human Rights in 1949; FDR Presidential Library & Museum 64-165

AI series: On AI ethics - influencing its use in the delivery of public good
AI
AI ethics

While AI algorithms can pose a number of challenges in terms of the size and sophistication of the algorithms, ethical issues can be the hardest and most important aspect to…

Olivia Varley-Winter
May 14, 2024

Stages involved in AI model development

AI series: Healthy datasets for optimised AI performance
AI
Data management
Data science education

Even “correct” data can cause all manner of mischief when fed indiscriminately into AI training and test data sets. Here the various issues that can arise and possible…

Fatemeh Torabi, Lewis Hotchkiss, Emma Squires, Prof. Simon E. Thompson and Prof. Ronan A. Lyons
May 7, 2024

A watercolour illustration in two strong colours showing the silhouettes of four people, two of whom have dogs on leads. They all cast shadows, and vary between realistic representations and those formed by representations of algorithms, data points or networks.

AI series: Generative AI models and the quest for human-level artificial intelligence
AI
Large language models
Machine learning
Deep neural networks

Diego Miranda-Saavedra explores some of the merits and limitations of modern machine learning models, and considers where these ‘intelligent’ systems might sit in the…

Diego Miranda-Saavedra
Apr 29, 2024

A schematic of a neural network depicted with circles connected with lines or arrows.

AI series: What is AI? Shedding light on the method and madness in these algorithms
AI
algorithms
statistics

AI has become a hot topic of debate but these discussions can circulate fears and fantasies more than a meaningful analysis of real world data. In this special issue we…

Anna Demming
Apr 22, 2024

Two bubbles overlapping on a black background.

What is data science? A closer look at science’s latest priority dispute
Algorithms
Data science education
Skills
Training

Two popular data science algorithms – naïve Bayes and eigen centrality – are used to examine the difference between data scientists, statisticians, and other occupations.

Jonathan Auerbach, David Kepplinger, and Nicholas Rios
Feb 19, 2024

Dark blue square with off-white rectangle at the bottom and small random white stars, and two white circles slightly off centre. Five black dots denote two eyes and three buttons on a snowman. Two brown lines look like arms. Two red rectangles form a hat. Orange triangle as a nose, with text reading Merry Christmas at the bottom.

Creating Christmas cards with R
R
Data visualisation

The programming language R is capable of creating a wide variety of geometric shapes that can be used to construct high quality graphics – including festive images. In this…

Nicola Rennie
Dec 12, 2023

Open lock and key. Photo by Basil James on Unsplash.

How to ‘open science’: A brief guide to principles and practices
Open science
Open source
Open data
Reproducible research

As a university lecturer, Isabel Sassoon made frequent use of open data resources for teaching and research. But it was only recently that she fully learned to embrace and…

Isabel Sassoon
Nov 6, 2023

DataScienceBites logo overlaid on a busy conference scene.

Heading to a conference this summer? Share your learnings here
Content ideas
Call for contributions
Events

Calling all summer conference delegates! If you’re heading to one of this year’s big statistics and data science events, consider writing about your favourite paper or…

Brian Tarran
Jul 17, 2023

New York road signs pointing left and right, both reading 'One Way'.

Choosing the right forecast
Forecasting
Prediction
Robustness

Practitioners often select forecast methods based on averages of scores from many probabilistic forecasts. But this is not without its difficulties. A recent paper by Bolin…

Brian King
Jul 13, 2023

Nine small images with schematic representations of differently shaped neural networks, a human hand making a different gesture is placed behind each network.

Trusted AI: translating AI ethics from theory into practice
AI ethics
Principles
Regulation

Data scientists can act as critical enablers of ethical AI when they have the right knowledge and toolkits at their disposal. Maxine Setiawan and Mira Pijselman review three…

Maxine Setiawan and Mira Pijselman
Jul 3, 2023

Portrait of Adolphe Quetelet, superimposed on a photo of blooming purple lilacs. Quetelet portrait from Wikimedia Commons, Public Domain. Lilacs photo by Arina Bogachyova/iStock.

A demonstration of the law of the flowering plants
Prediction
History of data science
Statistics

The day a flower blooms is one of the earliest phenomena studied with systematic data collection and analysis. The prediction rule developed nearly three centuries ago is…

Jonathan Auerbach
Apr 13, 2023

A shopping trolley being pushed around a supermarket. Photo by Marjan Blan | @marjanblan on Unsplash.

Using ‘basket complementarity’ to make product recommendations
Market basket analysis
Recommendation systems
Complementarity

Purchase suggestions – e.g., “if you are buying that, you might also want this” – are, to a large extent, informed by the concept of complementarity: that certain products…

Moinak Bhaduri
Mar 2, 2023

A researcher in a field shines a light on stylised waves of noise, as if attempting to identify or extract a signal. Image generated by DALL.E 2 from prompts provided by Real World Data Science.

Pulling patterns out of data with a graph
Data management
Dimension reduction
Graph theory

Large volumes of data are pouring in every day from scientific experiments, so much so that it is now commonplace to perform dimension reduction in order to reduce a large…

Andrew Saydjari
Jan 24, 2023

Photo taken from the backseat of a Lyft vehicle. Driver is seen in profile to the left of the picture. Phone is mounted on the dash and displays a maps app that is tracking the route to the passenger's destination. Photo by Paul Hanaoka on Unsplash.

Determining the best way to route drivers for ridesharing via reinforcement learning
A/B testing
Reinforcement learning
Statistics

A/B testing is often used to evaluate the impact of design ‘treatments’ — for example, are people who see advert A more likely to buy something than those who see advert B?…

Brian King
Dec 13, 2022
No matching items