Explore machine learning and voice recognition technologies for developers in this 59-minute Devoxx conference talk. Dive into the world of predictive capabilities and smart app development using Amazon's Machine Learning engine and Alexa Voice Service. Learn about key concepts such as reinforced learning, regression, validation, and overfitting. Follow along with a practical demo on model creation, feature engineering, and fine-tuning using AWS Analytics Console. Discover how to integrate deep learning, machine imagery, and graphical processing into your projects. Get hands-on experience with Alexa SDK, defining intents, and creating skills. Explore real-time prediction using Lambda functions, weather APIs, and logging requests. Gain insights into language support, sample utterances, and HTTP endpoint configuration for building engaging voice experiences.
Overview
Syllabus
Introduction
The London Underground
Customer approach
What is ML
Learning from Data
Reinforced Learning
Regression
Validation
Overfitting
Regularization
Model Creation
The Problem
Demo
Cuddle
Users
Data
Splitting Data
Features Engineering
AWS Analytics Console
Input Variables
Fine Tuning
Creating a Model
Models
Deep Learning
Machine Imager
Graphical Processing
Using Machine Learning with Voice
Alexa
SDK
Defining Intents
Code
Create a skill
Realtime prediction
lambda function
weather API
calling the ML system
logging requests
language support
intents
sample utterance
configuration
HTTP endpoint
Lambda
Request
Service Response
Echo Sim
Recap
Taught by
Devoxx
