Machine Learning Without the Hype

Explore the fundamentals of machine learning and artificial intelligence in this 52-minute Devoxx conference talk. Delve into the distinctions between AI, ML, and deep learning, and learn when to apply rule-based approaches versus machine learning techniques. Examine machine learning in the context of time-series data, understand the differences between supervised and unsupervised learning, and see a practical example using an actual dataset. Gain insights into various ML applications, including anomaly detection, visual inspections, and security event identification. Discover best practices for implementing machine learning, from data preparation and cleaning to model selection and evaluation. Analyze real-world scenarios like AWS outages and multi-cloud strategies, and learn to interpret correlations critically. By the end of this talk, acquire a solid foundation in machine learning concepts and their practical applications, free from hype and misconceptions.

Intro
Machine Learning Goes Viral
Fancy Technology
Gartner Hype Cycle
Machine Learning
General AI
Narrow AI
Automating Marketing
Chat Pods
AI API
Poncho
JetBot
What is Machine Learning
Machine Learning Definition
Linear Regression
Unsupervised ML
Reinforcements
Deep Learning
Deep Learning Example
Big Data
Machine Learning vs AI
Kegger contests
More productive
Finding patterns
Finding anomalies
Finding breakouts
Detecting anomalies
Detecting security events
Visual inspections
Rules
CPU intensive
Building the pipeline
Preparing the data
Data cleaning
Fastest ML algorithm
What is doable
What is the baseline
Multiple time series
GeoIP lookup
Default dashboard
Return codes
Top URLs
Operating Systems
Visualization
AWS outage
S3 outage
Multicloud strategy
Anomaly analysis
Combining multiple models
Correlation doesnt mean causation
Statistics class
Correlation
Future Predictions
Summary
Machine Learning Best Practices
Conclusion