Machine Learning Introduction @ Hack Gen Y

Machine learning Startup.ml Brings machine learning to startups Arshak

2 open source project vowpal wabbit apache aacumulo

text and image search spam detection speech recognition fraud detection intrusion detection in systems activitity recognition autonomous driving early epidemic detection THIS IS HOW THE CELL PHONES TURN ON TO VOICE

Bad uses of machine learning Banner ads, recommender systems Credit scores Google glass facial recognition Deep learning (feeding system for machine learning) Deep face (for facebook tag recognition)

Machine Learning Inputs ? Program (pramaters, instances) ? prediction

Predictions : Binary Classification, Various categorical divisions,

Regression problem How interested am i in a particular sport? Supervised learning technique (human taught it to machine with example data sets )

UNSUPERVISED: make 5 separate catagories, on your own, you decided you algorithm

Dimensional reduction preserve all these columns of data, but make it into a very few columns so a human can understand it

Inputs: Continuous: Income, age, time spent on page Categorical: state of residence, children, martial status Sparse : pages visted, collection of pixels; DON?T HAVE TOO Much info, only a few pieces of random data

Chris@gervang.com (spark core related work)

Daniel haaser Daniel@makeschool.com