Two talks all about the hottest topics in technology - edge computing, IoT and supervised machine learning. Gear up for an engrossing 3 hours of learning.
#1: Edge Computing & IoT with AWS Greengrass
AWS IoT allows the connected devices to interact with the cloud services and other devices. Edge computing brings the cloud capabilities closer to the devices. We shall talk about the challenges involved in the world of IoT and the role of edge computing in an IoT ecosystem. The talk will also cover how AWS Greengrass enables edge computing.
Speaker profile: Sreedevi is a Senior Software Engineer at Amazon. She has been in the software industry for over 15 years working on various embedded devices and middleware technologies. She has been associated with companies such as Nokia, Cisco in the past. At Amazon, she currently works on the Kindle E-Ink reader device. She is also an active volunteer at GHCI and plays the role of a track chair.
• Simplified architecture of an iot eco-system
• Things or sensors
b. Data model
c. Data aggregation & filtering
d. Offline operation
• AWS GREENGRASS
• Cloud interactions
Topics Covered: IoT, Edge Computing, AWS IoT, AWS Greengrass
#2: Supervised machine learning - A conceptual overview of methodology & techniques
Description: During the last 10 years, machine learning has rapidly evolved from being a technology largely confined to applied research and a small number of niche applications to a buzzword in the IT circles and is now a mainstream technology with myriad real-world applications, many of which would have been considered practically infeasible not too long ago. Interestingly, machine learning in itself continues to be in a state of constant flux as new learning and knowledge-representation techniques emerge and change the state-of-the-art at a pace not seen in most mainstream technologies. Yet, some of the basic concepts and methodologies associated with the older techniques continue to be relevant to the newer techniques too. This talk will present an overview of such key concepts and standard algorithms associated with supervised machine learning, and some important points to be considered when applying machine learning to solve a problem.
Speaker Profile: Sriraghavendra is presently an applied scientist in the e-book content quality team. He has over 10 years of experience as a software engineer and has worked on design and implementation of backend services, workflow-based applications and algorithmic solutions for various teams in the Kindle content organization. Though he is a relatively new entrant to applied machine learning at work, his introduction to machine learning pre-dates his career at Amazon. He is amazed to find that some aspects of machine learning have come a full circle since the time he was first introduced to that field.
* Introduction to machine learning
* The model building process for supervised learning
* Overview of some classification methods - tree-based classifiers, linear models, ensemble classifiers
* Model complexity, parameters vs. hyper-parameters, importance of feature engineering, other factors to be considered when building a machine-learnt solution
* [Subject to availability of time] A shallow overview of deep learning. Deep learning vs. statistical inference methods
Pre-requisites : Basic knowledge of algorithmic problem-solving to be able to appreciate machine learning techniques. A basic understanding of probability and statistics will be helpful