MICROSOFT AZURE MACHINE LEARNING STUDIO IN REAL WORLD

  • Slides: 15
Download presentation
MICROSOFT AZURE MACHINE LEARNING STUDIO IN REAL WORLD SPORTS Global Azure Bootcamp Zagreb, Croatia

MICROSOFT AZURE MACHINE LEARNING STUDIO IN REAL WORLD SPORTS Global Azure Bootcamp Zagreb, Croatia 20180421

About us Darko Katović Kineziološki Fakultet KIF Miljenko Cvjetko - moljac Microsoft (Xamarin Inc.

About us Darko Katović Kineziološki Fakultet KIF Miljenko Cvjetko - moljac Microsoft (Xamarin Inc. )

Intro – Motivation - Professional ________ (fill in some hype word here) Options (Machine

Intro – Motivation - Professional ________ (fill in some hype word here) Options (Machine Learning, Data Science, Deep Learning, …) Darko __________ does for living (teaching) moljac Works for Xamarin (Microsoft) on tools for apps, so logically writing apps makes him happy Xamarin does tools for Mobile apps w/o backend are toys

Intro – Motivation - Personal

Intro – Motivation - Personal

Intro – What is ______? Data Science Data wrangling (munging), retrieval + storage Statistics

Intro – What is ______? Data Science Data wrangling (munging), retrieval + storage Statistics Big Data Machine Learning Deep Learning Soft Computing Fuzzy Data mining & machine learning Big data Genetic Data Wrangling/Munging Data Storage Data Manipulation Statistics

Intro – What is ______? Detailed Machine Learning == Data Mining ? ? Data

Intro – What is ______? Detailed Machine Learning == Data Mining ? ? Data Acquisition Data Analysis (exploration) Pattern / Rule finding Modelling – Model building Predictions / Decisions

Intro – How? Business Problem Business problem Sales must reach == qualify for Olympics

Intro – How? Business Problem Business problem Sales must reach == qualify for Olympics Data Solutions - optimizations Data Sports – amount of data sufficient ? ? ? Models (read Math) Models Complex (high number of parameters) Non linear

Intro – How? Detailed PROCESS Business Problem Goal: Win, Olympics Optimizations (minimal energy or

Intro – How? Detailed PROCESS Business Problem Goal: Win, Olympics Optimizations (minimal energy or time) Optimizations (no injuries, least effort / cost, ASAP, Data Collect data (activities, tests, competition) Models (Math) Complex high number of parameters) Non linear Data: Activities, Tests, Competition Models (Math) Complex Fitness (Physiology), Motorics, Psychological y = ax + b

Data Analysis and Modelling - Model Definition Data Collect Exact (if possible) Train Store

Data Analysis and Modelling - Model Definition Data Collect Exact (if possible) Train Store Validate (score) Analyze Use it – Explore or Deploy Extend/Update it

Demo – Regression. Linear. Weight. Control

Demo – Regression. Linear. Weight. Control

Demo - Correlation. Basketball Machine Learning Studio help in real world applications

Demo - Correlation. Basketball Machine Learning Studio help in real world applications

Demo Machine Learning Studio help in real world applications

Demo Machine Learning Studio help in real world applications

Where are we? Data In development – application suite (xplat horizontally and vertically) Io.

Where are we? Data In development – application suite (xplat horizontally and vertically) Io. T Mobile Desktop Server (ASP. net Core) Collecting (from web and with the apps) Models, Modelling Testing, playing

Outro Machine Learning Studio help in real world applications

Outro Machine Learning Studio help in real world applications

Q&A Shoot

Q&A Shoot