Topological Data Analysis - New Perspectives on Machine Learning - by Jesse Johnson

Explore topological data analysis and its applications in machine learning through this insightful ODSC Meetup talk. Delve into how geometric patterns and shapes in data collections can be leveraged to understand machine learning algorithms. Learn about higher-dimensional abstract geometry and topology concepts and their role in uncovering data structures. Discover how Jesse Johnson, a former math professor specializing in abstract geometry and topology, applies his expertise to develop new algorithms and make machine learning concepts accessible to non-experts. Gain valuable insights into mathematical models, probability distributions, classification, clustering, and geometric structures in various data types including words, sentences, sounds, and pictures. Understand the importance of geometry in data analysis, from simple to complex structures, and explore advanced concepts like persistent homology, learning curves, and local analysis.

Introduction
Mathematical Models
Probability Distribution
Analysis
Classification
Clustering
FourDimensional Data
FourDimensional Geometry
Iris Data
Categories
Geometricity
Words
Sentences
Engrams
Sound
Pictures
Kmeans
Results
Why Geometry
Simple Structures
Geometric Structures
Complex Structures
Persistent Homology
Learning Curves
Scaling
Sampling
Local analysis