Data Science for the Internet of Things (IoT)
Be the first to know about future course dates - click the 'Notify me' button on the right of this page.
This unique course aims to create a new breed of engineer - those with a background in IoT and knowledge of machine learning, AI, cloud and robotics.
The Data Science for IoT course aims to equip you with the skills to solve problems, providing you with a toolkit (code) and templates. It's for developers who want to be data scientists with an emphasis on IoT.
The course explores problem solving for IoT analytics via the following topics:
- Concepts: principles/foundations
- Product development for IoT (with an emphasis on analytics)
- Data science
- Artificial intelligence (AI)
- TensorFlow and Keras
- Deploying AI models to scale (e.g. via Kubernetes)
- IoT Verticals
- Time series
- Deep learning
- Real time including LSTMs and streaming
- NoSQL databases for IoT
- IoT data visualization
- Industrial IoT
- Robotics and drones
- Edge analytics
- Complex event processing
- Innovation in IoT
- Methodology – putting it all together
About the course and its aims:
- The course analyses problem solving for IoT analytics.
- The unique considerations for IoT data (e.g. time series data) are investigated.
- The course covers programming so participants will need to be familiar with some programming languages - but we do not expect familiarity in a specific language. The primary programming language of the course is Python (specifically TensorFlow and Keras).
- We use Spark for big data.
- The course needs an understanding of maths. We cover maths and statistics foundations as needed.
- Where possible, we use IoT datasets. We cover handling large-scale IoT datasets.
- We focus on skills based/commercial products. This is not an academic course.
- The course also includes an industry programme. The industry programme will be based on use cases incorporating IoT analytics methodology.
We aim to equip you with skills such as TensorFlow, Keras, Nvidia etc., which can be used outside of IoT applications.
The course takes a problem solving approach and uses specific case studies from industry. Participants are expected to have a mind-set of exploration and to study and learn beyond the class material itself (depending on their existing familiarity with the subject matter).
The course is based on a perspective of both AI and machine learning. AI is driven by deep learning algorithms. Deep learning is a wider case of machine learning based on automatic feature detection. IoT primarily involves data in time series formats (using AI algorithms like recurrent neural networks and long short-term memory (LSTMs)) and image-based data (using convolutional neural networks).
A limited number of participants ensures that all those taking this course gain the maximum possible value.
All participants finishing the Data Science for IoT course will receive a University of Oxford certificate showing that they have completed the course (see sample and details of requirements further down this page).
Time commitment for the course is 2 - 3 hours face-to-face in Oxford on Saturdays (usually starting at 10:30) and 1 - 2 hours online each week on Tuesdays (usually starting at 19:00). We recommend you allow around 10 - 12 hours study time per week (plus the hours above). There is a minimum attendance requirement of 75%.
You will be fully supported by the tutor who will be available during the week to answer questions.
The tutor will also offer a number of one-to-one 'surgery sessions' during the course.
All course participants will receive a Nvidia Jetson TX2 developer kit worth £441.00 (included in course fee).
All course participants will receive copies of these books (included in course fee):
Deep Learning with Python
(2017, Manning Publications)
Deep Learning with Python
Dr Jason Brownlee
(2016, Machine Learning Mastery)
Python Machine Learning
(2015, Packt Publishing)
Big Data Analytics with Spark: A Practitioner's Guide to Using Spark for Large Scale Data Analysis
Data Science for Internet of Things
Ajit Jaokar and Jean-Jacques Bernard
(forthcoming edition Nov 2017, Futuretext)
Mathematical Foundations for Data Science (awaiting confirmation of publication)
Travel to Oxford
The course will be held at our Rewley House site, which is a short walk from Oxford bus and train stations. Oxford is approximately 55 minutes by train from London Paddington and London Marylebone train stations. We are also about one hour from Birmingham. See the National Rail Enquiries website for train information.
Week Zero: Onboarding platform, introductions etc.
Week One: Foundations of Data Science for IoT
- Explore the application of predictive learning algorithms to IoT datasets in detail.
- Discuss the modalities where we could apply IoT analytics (Streaming, Edge etc.).
- Understand data science for IoT from a problem solving perspective.
- Understand IoT platforms and their strengths and drawbacks
- Explore where analytics fits into the overall Product development for an IoT based application
- IoT verticals
Week Two: Programming Foundations (part one) - The Python Ecosystem
In this session, we explore Python from a data science perspective using the book Python Machine Learning by Sebastian Raschka
Week Three: Programming Foundations (part two) - Understanding Tensorflow and Keras
Building on Week Two, we explore TensorFlow and Keras in detail using the books
Week Four: End to End IoT Deployment
This session covers an end to end IoT deployment covering sensors, analytics and visualization for devices using Microsoft Azure.
Week Five: Deploying Machine Learning and AI Models in Production
In this session, we explore the deployment of machine learning and AI models in production using two models: TensorFlow, Keras with Kubernetes and h2o.ai
Week Six: Time Series Analysis
Most IoT data is in time series format. In this session, we explore time series analysis including recurrent neural networks, LSTMs and multivariate time series.
Week Seven: Understanding and Managing Data and Models
Understanding and pre-processing data is a key part of data science. For IoT, we need two types of data, namely time series and image-based data. This session is based on understanding the transformations and models needed for IoT data with an emphasis on time series data and image based data.
Week Eight: Big Data and IoT - Spark Streaming and NoSQL Databases
In this session, we explore the impact of big data on data science for IoT. This includes perspectives like streaming (Spark) and NoSQL databases.
Book used for this section: Big Data Analytics with Spark: A Practitioner's Guide to Using Spark for Large Scale Data Analysis by Mohammed Guller.
Week Nine: Robotics, Drones etc. with Nvidia using AI and Deep Learning
With cameras as sensors, AI and deep learning technologies play a key role in a number of emerging areas. In this session, we use Nvidia platforms to understand the application of deep learning algorithms to a range of emerging technologies including drones, robotics etc. We primarily address video and image recognition based problems.
Week Ten: Industrial IoT
This session discusses various aspects of Industrial IoT including deployment of complex event processing, Industrial Internet of Things (IIoT) Platforms, platforms such as Predix, physics based modelling etc.
Week Eleven: Innovation / Latest Developments for Data Science for IoT
This session covers a range of innovations and latest developments with respect to data science for IoT including GPU databases, Voice interfaces for sensors, Blockchain, AI at chipset level, Tensorflow for embedded devices etc.
Week twelve: Industry Use Cases
Industry use cases and meeting with invited experts - spanning all the sections above, a discussion and collaboration with invited industry experts from Ocado, Barclays, Google, Microsoft, GSMA, Logtrust and others.
Note that tutors and content may be subject to minor revision during the course development process.
Participants who satisfy the course requirements will receive a Certificate of Attendance. The sample shown is an illustration only and the wording will reflect the course and dates attended.
To receive a certificate at the end of the course you will need to:
- Achieve a minimum attendance at the Oxford classroom sessions of 75%.
- Answer all the weekly learning quizzes (these are short quizzes designed to ensure you have understood the material in each unit).
- Complete the short exercises that you are given.
Standard course fee: £3995.00
Lead Tutor and Course Developer
Author and Big Data/IoT/Telecoms Specialist
Ajit's work work spans research, entrepreneurship and academia relating to artificial intelligence (AI), the internet of things (IoT), predictive analytics and mobility.
Ajit works as a Data Scientist (fintech and IoT). He is also the Director of the newly founded AI/Deep Learning Labs for Future Cities at UPM (Universidad Politécnica de Madrid). Ajit is also currently the Research Data Scientist at a fintech firm in London.
Ajit publishes extensively on KDnuggets and Data Science Central and his book, Data Science for Internet of Things, is included as a course book at Stanford University.
He was recently included in top 16 influencers (Data Science Central), Top 100 blogs (KDnuggets), Top 50 (IoT central), No 19 among top 50 twitter IoT influencers (IoT Institute).
Ajit has been involved with various mobile, telecoms and IoT projects since 1999 including strategic analysis, development, research, consultancy and project management.
In 2009, he was nominated to the World Economic Forum’s ‘Future of the Internet’ council. In 2016 he was involved in a WEF council for systemic risk (IoT, drones etc.). He has worked with cities like Amsterdam and Liverpool on Smart City projects in mayoral level advisory roles. Ajit has been involved in IoT-based roles for the webinos project (Fp7 project). In May 2005 he founded the OpenGardens blog, which is widely respected in the industry. He has spoken at Mobile World Congress (4 times), CTIA, CEBIT, Web 2.0 expo, The European Parliament, Stanford University, MIT Sloan, Fraunhofer FOKUS; University of St Gallen. He has also been involved in transatlantic technology policy discussions.
Ajit is also passionate about teaching data science to young people through space exploration working with Ardusat.
Director of Studies
Associate Professor of Data Science, Department for Continuing Education, University of Oxford
Dr Cezar Ionescu is Associate Professor of Data Science at the Department for Continuing Education and Director of Studies in Computing and Mathematics.
His research interests include functional programming, dependently-typed programming, scientific computing, computing in schools, algorithmic thinking, synthetic populations.
Director, Insight and Customer Strategy, Oracle
JJ Bernard has an MSc in Engineering from the Ecole Centrale de Marseille and an MBA from the University of Cambridge. He is also a certified Lean Six Sigma Black Belt.
He currently works at Oracle as a strategy consultant, advising client on their technology investment with regards to their business strategy.
Previously, he has worked as a project manager for Orphoz, a subsidiary of McKinsey & Company, focusing on defining and executing transformation projects. He has worked in diverse environments such as manufacturing, FMCG, public sector, steelmaking, mining, healthcare and automative.
He has also worked for Accenture, both on business and IT transformation projects, for the financial services industry and the public sector.
Founder, Seven Symbols
Barend Botha is a UK-based consultant working predominantly in the field of data visualisation.
He is particularly interested in the growing future potential and roles that data analytics and visualisation will play across IoT verticals in in combination with Artificial Intelligence and the resulting products and insights for business and consumers alike.
He draws upon past experience in research, design, development, management and marketing across various domains and disciplines.
Jo-fai (Joe) Chow, Data Scientist, H2O.ai
Before joining H2O, Joe was in the business intelligence team at Virgin Media in the UK where he developed data products to enable quick and smart business decisions.
He also worked remotely for Domino Data Lab in US as a data science evangelist promoting products via blogging and giving talks at meetups. Joe has a background in water engineering.
Before his data science journey, he was an EngD research engineer at STREAM Industrial Doctorate Centre working on machine learning techniques for drainage design optimisation.
Prior to that, he was an asset management consultant specialised in data mining and constrained optimisation for the utilities sector in the UK and abroad.
Joe also holds an MSc in Environmental Management and a BEng in Civil Engineering.
Chief Technology Officer, Ocado
Paul Clarke is Chief Technology Officer at Ocado, the world's largest online-only grocery retailer.
Paul joined Ocado in 2006. After establishing new teams for Simulation and Mobile development, Paul then co-wrote the first of Ocado’s award winning mobile apps. In his current role, Paul heads up Ocado Technology, whose 950+ software engineers and other IT specialists are responsible for building all the software and IT infrastructure that powers Ocado, and now Morrisons’ online grocery business too.
Paul read Physics at St John's College, Oxford before then entering the computer industry. He has worked in software engineering, consultancy, interim management and a number of software start-ups.
In what little spare time he has alongside his work and family, Paul loves to invent and build stuff, design PCBs, write software and generally tinker.
Principal Software Engineer, Microsoft
Since joining Microsoft in 1994, Paul Foster has worked across a wide range of sectors and customers, providing a mix of technical and strategic guidance around the creative use of technology in relation to their business needs.
Paul is currently a principal software engineer at Microsoft. As an established public speaker across Europe and having spent a considerable amount of time working on the cutting edge of technology providing leadership and inspiration on topics like Smart Devices, Cloud Computing, Education and App Development.
Paul is currently focusing on the building of next generation sensor webs which automate the gathering of data from disparate sources, and how to enable the creative analysis of this data to start a new era of perception.
For a short time Paul was a member of a high-wire flying trapeze circus troupe, is a keen roboticist and an international marathon runner.
Imperial College London; Organiser, Google Developer Group (GDG) Cloud London
Surya is pursuing an MEng in EEE from Imperial College London, specialising in Machine Learning and Embedded Systems. He started his journey in data science at a biotechnology startup, where he worked on characterising hand tremors of Parkinson’s patients to tune the product’s control algorithm (which stabilises the tremors). He spent last summer working on implementing neural network based applications on Android.
He also leads workshops and events at the Cloud Google Developer Group in London, with a specific focus on Machine Learning frameworks, such as Tensorflow.
Surya can usually be found tinkering with embedded hardware and software in the Robotics Lab at Imperial College.
Director & Technical Consultant, Catalyst Computing Services
Peter Marriott is an industry practitioner with over 25 years database and software development experience on production systems. The main focus of his career has been working with data, working on business systems to make the data useful and available in a performant cost-effective way.
13 years ago he founded the consultancy Catalyst Computing Services, working with a wide range of clients in a variety of market sectors. Their first IoT project was in 2006. Peter has been involved working throughout the whole project development cycle of IoT: from prototyping proof of concepts; designing systems architecture; development; deployment and training of support staff.
Peter runs training courses in cloud computing and database technologies for clients, gives career talks for undergraduates and speaks at various IT user groups.
Head of Advanced Data Analytics, Barclays
Harry Powell runs the Advanced Data Analytics team at Barclays which pioneered the use of Apache Spark in financial services businesses in London.
Co-founder of Blesh
Devrim Sonmez is regarded as a high performing executive and entrepreneur with over 19 years of experience at intersection of digital and technology with various industries, over 10 years of Executive Board experience and taking a strategic role in securing M&A deals worth over $170 million in technology.
In 2013, he became the CEO and co-founder of Blesh Inc. which is an award winning and innovative start up in IoT space developing IoT platforms and solutions through its own research and development. It is among the top Global 5 Mobile Proximity Solution providers and is considered the first global beacon solution provider of Google Physical Web technology since 2014 with exports more than 30 countries.
He is still the Cofounder and the Board Member of Blesh. Alongside these positions, Mr. Sonmez has sat on the board of several more companies and associations including IOT.ATL a program of US Metro Atlanta Chamber that intends to boost the region’s tech sector in the burgeoning industry of Internet of Things products and software, the highly prestigious TUBISAD (Turkish Informatics Industrialists Association) and English Ninjas (An innovative mobile startup for English practicing), where his influence plays an integral part in grow and success.
Devrim holds a Mathematics Bachelor of Science from the leading Turkish university in Ankara, the Middle Eastern Technical University graduating in 1998 and a Master of Business Administration from the Koc University in Istanbul, graduating in 2005.
CTO Co-founder, "Steath" Medical Startup
Petteri is a multidisciplinary entrepreneurial scientist with a MSc in electrical engineering from Aalto University, and a PhD in visual neurosciences from University of Lyon. He is currently working on his own early-stage AI-based healthcare diagnostics and disease management startup. Previously, Petteri had worked for a proptech startup Cubicasa developing deep learning computer vision models for real estate point clouds, and for eyecare startup Visulytix designing deep learning clinical classification models. Before transitioning to industry, he did two post-docs in North America involving projects on two-photon microscopy, brain stimulation, image segmentation, electronics design and electrophysiology.
Terms and conditions
Terms and conditions for applicants and students on this course
Sources of funding
Information on financial support
Follow tweets by @DataScience_IoT