Before giving an overview of some the key content that will be covered in A.I.Camp, I should introduce myself and some of my experiences.
I am a software engineer at Kainos working with the applied innovation team, my role involves investigating emerging technologies and communicating the value of them to various clients and internally within Kainos. I have spent time within the team knee-deep in machine learning, and have been involved in various projects using this technology, such as automated pill counting in the medical domain. I am one of the organisers of A.I.Camp with a focus on content creation, delivery of training materials and other logistics.
Enough about me, now the good stuff!
A.I.Camp is a brand-spanking new initiative created by Kainos with the goal of taking university students who have a latent interest in AI and providing a two week crash course on creating machine learning driven processes.
If you are new to the topic you might ask, what the heck is machine learning? well…
Machine learning is a type of artificial intelligence that provides computers with the ability to learn without being explicitly programmed.
This is the topic in a nutshell, however machine learning is vast ocean of ideas so for A.I.Camp it is crucial we focus on the concepts that we feel have the most value now and in the near future. Lets dive right in.
Computer vision of a subset of machine learning which focuses on giving computers the ability to gain an understanding of the content in digital images or videos. This includes performing certain tasks such as object recognition, image manipulation, event detection and others.
How do these tasks relate to the real world? Below are a few applications of computer vision in our day to day lives:
- Self driving cars — computer vision techniques help detect obstacles
- Snapchat filters — the app needs to know how to find your face somehow
- The Google photo app uses this technique to let you ‘search’ your images.
You can also do some cool things like this!
Without data, machine learning is nothing. As part of A.I.Camp a consistent theme will be the use of a dataset in order to establish the ‘intelligence’ of our machine learning processes and enable informed decision making. This data can take multiple forms such as images, text and even audio.
Throughout the camp’s duration students will learn how to:
- Extract features from data — for an image, this would include edges or colour
- Clean datasets — preprocessing such as removing bias and handling missing values
- Identify what makes up a ‘good’ and a ‘bad’ dataset and from this determine how easy it will be to work with it effectively.
This is a term you might have come across before, but what actually are they? well according to Google, chat bots are:
A computer program designed to simulate conversation with human users, especially over the Internet.
This sums up the topic quite well, at least from our perspective for A.I.Camp. Intelligent chat bots have the potential to transform user experience on the web, we are already seeing bots that can be used for symptom diagnosis and customer service sneaking into our day to day lives through streams such as social media (Facebook / Twitter bots etc.)
- Maturity of services — in recent years the ability to create chat bots has been aided by the established of key services such as Microsoft’s bot framework and IBM’s Watson conversation service. These services make it easier to not only create the chat bot but integrate with other platforms such as social media, Skype and others
- Advancements in machine learning — Connecting bots to a machine learning process in the backend to make decisions, remember conversational context and adapt based on user input is where the real value of chat bots lie. Now more the ever, machine learning is at its most accessible and developer friendly to compliment this.
Software Engineering With Machine Learning
Machine learning is a powerful tool, however in certain use cases it will only help solve part of the problem – especially in a software development context. Therefore it is crucial to not only learn about machine learning in isolation but the software components that surround it, such as a user interface.
Soft Skill Development / Hackathon
An inherent part of A.I.Camp will be the development of technical skills, this is an expected side effect of building machine learning driven solutions. A secondary goal of A.I.Camp is the development of soft skills, such as:
- Problem solving – Students will be given the autonomy to tackle challenges on their own and through using their own critical thinking skills.
- Communication – Being able to convey an idea or solution in a succinct manner with an overarching story is crucial, particular in the concluding Hackathon (more on this later)
- Team work – Not only will collaboration be encouraged during the day to day activities of A.I.Camp, students at times will have to form into groups to solve a particular problem.
At the end of the two week camp the students will form groups and take part in a Hackathon with the goal of devising a solution that tackles a specific theme, creates something innovative or solves a problem. The soft skills they have developed over the course of the camp will be crucial in order to work well as a team and present a compelling story that reinforces their solution alongside a presentation.
A.I.Camp is doing something ambitious, taking a series of complex topics within machine learning and distilling the knowledge into actionable skills that can translate across multiple domains. This is where the value of the camp lies; building a fundamental knowledge base for students to equip them with the skills needed to apply machine learning towards real-world use cases.
Follow the twitter account @KainosAcademy