How to build a neural network in under 5 minutes using Brain.js

Artificial neural networks are one of the many approaches to Machine Learning (ML). It is based on the neural network of the human body where each neuron keeps learning and interconnects with other neurons to perform different actions based on different responses. We are currently surrounded by several neural networks used in various fields to solve different kinds of problems. This being the case, for most of us building and training a neural network still remains a mystery.

Come attend this session, where I demystify neural networks and show how easy it is to build your own neural network in under 5 minutes using Brain.js. I will first cover the basics of AI, ML and Deep learning. Then, I will walk you through a step by step process of building a neural network in a live demo session. Discuss how this is applied to the field of software development and testing. This session will prove to you that, there is no need to be a data scientist to train your own models. Anyone just interested to learn about AI can get started on this journey. Both technical and non-technical people will be able to build their own models after this session.

This session will prove to you that, there is no need to be a data scientist to train your own models. Anyone just interested to learn about AI can get started on this journey using the power of open source frameworks. In this case it is Brain.js.

 
 

Outline/Structure of the Demonstration

Part 1- How AI based systems train and work - 15 mins

  • Discuss basics of AI
  • Show a demo of how AI based training works using open source application - Google Draw and talktotransformer
  • Show quickly Brain.js and how deep neural network training works
  • Debunk some of the myths surrounding AI

PART 2 - Building your own neural network with Brain.js - 20 mins

  • A demonstration of building an neural network using Brain.js
  • How this training is applied to the field of software development/testing
  • Address challenges interacting with AI based systems

PART 3 - The future of AI based technologies - 10 mins

Learning Outcome

Takeaways

  • Learn with live demos how AI model training happens
  • Learn how to build your own neural network with live demos using Brain.js
  • Learn about the future of AI based technologies

Target Audience

Anyone willing to learn about open source AI libraries

Prerequisites for Attendees

Nothing.

schedule Submitted 4 months ago

Public Feedback

comment Suggest improvements to the Speaker
  • Robin Gupta
    By Robin Gupta  ~  3 months ago
    reply Reply

    Can you please add relevant details in regards to the below points:

    • How is the topic at hand overlapping with the themes of testing/automation/selenium?
    • Is Javascript/Machine learning a pre-requisite for the session?
    • Can the concepts taught here, be ported to other libraries such as fast.ai, tensorflow, DL4J etc. ?
    • Raj Subrameyer
      By Raj Subrameyer  ~  3 months ago
      reply Reply

      Great questions. Here are the answers-

      • How is the topic at hand overlapping with the themes of testing/automation/selenium?

      Currently, most of the companies using AI for Test Automation are using different neural network algorithms to figure out different patterns regarding tests, such as -  the most commonly run test scenarios, which tests are flaky, how much time a user spends on a particular module of an application, the most used test data combination, extraction of attributes of each and every element on a page to adapt to changes in the GUI and so on..

      Not many people are aware of how the AI tools can figure out these patterns. This is what I will discuss in this talk. I am going to take an open source library called Brain.js. I am going to use simple datasets and show how a neural network is trained and what kind of predictions it makes.

      My talk will help people understand the internals of how AI model training works and how it is applied in the domain of test automation.

      • Is Javascript/Machine learning a pre-requisite for the session?

      Nope. There is no pre-requisite for my talk. Attendees can just follow along with my live demo. Before, I show live neural network training, I cover some basics on AI, Deep Learning and ML, So people get some context on what they actually mean and how they work before I get into the live demo. 

      Thus, my talk will set some context on AI, ML, Deep Learning, then get into the neural network training and finally finish off with future of test automation and AI based technologies.

      • Can the concepts taught here, be ported to other libraries such as fast.ai, tensorflow, DL4J etc. ?

      Yes, the concepts shown in my live demo are the exact things used in any library. The concepts include-

      • strategy to split test datasets (input/output combination)
      • applying more hidden layers
      • using an activation function (sigmoid, tangent, RELU or ELU)
      • Overfitting and underfitting in AI model training
      • Robin Gupta
        By Robin Gupta  ~  3 months ago
        reply Reply

        Thanks for the replies.

        • Raj Subrameyer
          By Raj Subrameyer  ~  2 months ago
          reply Reply

          Hi Robin,

          I added more details to the outline of this talk with time splits and added more clarity on the key takeaways. Just wanted to let you know.

          • Robin Gupta
            By Robin Gupta  ~  2 months ago
            reply Reply

            Sure thing.

             

            THanks..


  • Liked Shweta Sharma
    keyboard_arrow_down

    Shweta Sharma / Nikita Jain - Accessibility testing 101

    45 Mins
    Talk
    Beginner

    "This world is such a beautiful place to live in." If you can read the first sentence without any screen readers or assistance, you're privileged. As technologists, shouldn’t we be more empathetic towards differently-abled people and make all parts of our website accessible to them? In my humble opinion, the true power of technology can be identified when it reaches out to people of all kinds having different physical or psychological challenges. We not only legally bound to provide Accessibility but also it should be considered as our moral responsibility.

    As testers, we have a wonderful opportunity to contribute to Accessibility by ensuring that the site is accessible in many different ways. Although it is impossible to identify all the issues that exist around Accessibility in the world, we are lucky enough to still understand a majority of them. With this understanding, there have been many measures taken in order to make your site accessible. But, don’t forget - we are QA engineers. We got to ensure that the site is accessible as per the standards set by WCAG 2.0 (AA) by testing for accessibility using various tools and techniques.

  • Liked Ivan Krutov
    keyboard_arrow_down

    Ivan Krutov - Master-class: Bulletproof Selenium cluster

    Ivan Krutov
    Ivan Krutov
    Developer
    Aerokube
    schedule 4 months ago
    Sold Out!
    45 Mins
    Demonstration
    Beginner

    Browser tests are known to be the flakiest ones. This is partly because browser infrastructure is complicated to maintain. But the second reason is – mainstream browser automation tools such as Selenium server are far from being efficient.

    During my previous talks I was speaking about Selenoid - a truly efficient replacement of the standard Selenium server. This year I would like to do a live demonstration how to organize a fault-tolerant and easily scalable Selenium cluster using virtual machines in the cloud. I will start by setting up Selenoid and show its powerful features like video recording, live tests debugging, manual testing and many more. Then I will configure Selenoid to send logs and recorded videos to S3-compatible storage. Finally we will run a Ggr load balancer instance allowing to use all running Selenoid nodes and organize a single entry point to the cluster.

  • Liked Raj Subrameyer
    keyboard_arrow_down

    Raj Subrameyer - Viewing Software through the lens of AI - The impact to software development and testing

    45 Mins
    Talk
    Beginner

    A recent study conducted by Gartner, shows that by 2020, AI will be pervasive in almost all software products and services. The major highlight of the study was how our skills as engineers would have to adapt accordingly. The current roles in companies are going to change significantly and we need to be prepared for it. There are areas of healthcare, finance, retail, customer analytics, and manufacturing that have already started seeing positive effects of using AI, many of which we would have never thought possible just a few years ago. At the same time, we have downsides of using AI, such as how it can easily impact our data and privacy if we do not pay attention to how the AI models are trained. The working of AI is a black box-- we do not control or understand how the algorithm forms different relationships and makes decisions -- we just provide different training datasets and monitor the learning/progress. We are trying to make predictions on future values based on learning from past examples, or trying to discover different patterns from datasets. This being the case how do we approach these systems?

    In this talk, I will discuss the basics of how AI based systems work, what are the key factors to consider when developing and testing these systems and how to come up with an enterprise wide strategy. I will provide insights into what skill sets would be needed by developers and testers to co-exist in the world of AI. Finally, I will address some common questions and concerns surrounding AI and also show what are the tools available for doing AI based development and testing. This talk will help anyone get a basic understanding of developing and testing AI based systems and make us aware of the key factors to consider to ensure these systems do not have a negative impact on humans.

    Key Takeaways-

    • Learn with live demos how AI model training happens
    • Learn how AI is used in software development and testing
    • Learn what AI based tool and frameworks are available for software development and testing
    • Learn key factors to ensure these systems are safe for humans

  • Liked Raj Subrameyer
    keyboard_arrow_down

    Raj Subrameyer - Reaching your highest potential in a chaotic environment

    45 Mins
    Talk
    Beginner

    While working on our daily tasks in agile teams, we quite often have this feeling where we are working on multiple tasks all day long and at the end of day when we review our work, we realize we haven’t accomplished anything concrete. The main reason for this is, our work environments are filled with distractions from having unnecessary and unproductive meetings to people constantly checking their messages on phones, emails and slack channels. As a result, we feel demotivated, less productive and burnt out by the end of the day. Burnout is a serious issue and is attributed to a lot of mental and health issues. I was one such person who after working around the clock and helping people 24/7 ended up getting admitted to the hospital. How do we prevent this repetitive cycle from happening on a daily basis? This is where mindfulness and productivity hacks can help.

    In this talk, I will share my real life experiences from experimenting with different mindfulness and productivity hacks which ended up helping me get more focus, regain my mental and physical health and most of all made me fall in love with technology all over again. Some of these hacks include doing deep work, different tips and tricks to prioritize tasks, deal with emails, make meetings more productive, effective engagement with remote teams and much more. Come join my journey and become the best version of yourself.