Scribd
Upload
20 views10 pages

Applications of Data Mining & Computational Pattern Recognition

This document discusses applications of data mining and computational pattern recognition. It provides examples of how Netflix used these techniques in a video recommendation contest to improve accuracy by 7%. It also mentions how Google's use of these techniques helped it dominate web searches. Finally, it outlines different types of pattern recognition systems, including supervised and unsupervised learning, and some common algorithms and weaknesses in generalizing to new patterns.

Uploaded by

Ayaz
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
20 views10 pages

Applications of Data Mining & Computational Pattern Recognition

This document discusses applications of data mining and computational pattern recognition. It provides examples of how Netflix used these techniques in a video recommendation contest to improve accuracy by 7%. It also mentions how Google's use of these techniques helped it dominate web searches. Finally, it outlines different types of pattern recognition systems, including supervised and unsupervised learning, and some common algorithms and weaknesses in generalizing to new patterns.

Uploaded by

Ayaz
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
You are on page 1/ 10

Applications of Data Mining & Computational Pattern Recognition

Netflix
Grand Challenge (2006)
Objective: Enhanced accuracy of video recommendation system by 10% Prize: $1 million ($50k to the progress leader for the duration of the contest) Status: 7% overall improvement as of April 2008

Google
Google It! Innovation: novel reference influenced ranking system spawns the field of Collective Intelligence Payoff: In 2004, only 6 years from its inception, Google handled 85% of online web searches!
Its founders are now among the top 10 richest people in the world!

Others
Biotech Investigation & Fraud Detection Image recognition & Machine Vision Product Marketing Supply Chain Optimization Infrastructure management Financial market analysis National Security

Technological Advantage
Collective Intelligence The combining of behavior, preferences, or ideas of a group of people to create novel insights. Pattern Recognition The act of taking in raw data and taking an action based on the category of the data

Pattern Recognition Systems


Sensor: Data generator/gatherer Feature Extractor: A symbolic or numerical transformation/representation of the sensor data Classifier: A description scheme/ mechanism for describing observations based on extracted features

Pattern Recognition Systems


Learning paradigm: Supervised: Labeled training data provided ex: Gene prediction Speech & handwriting recognition Unsupervised: Classes based on pattern regularities ex: Google pagerank eHarmony

Pattern Recognition Systems


Algorithms Bayesian classifiers Decision Trees Neural Networks Support Vector Machines Self Organizing Maps

Pattern Recognition Systems


Primary weakness All data mining / pattern recognition algorithms vary in their ability to generalize over large sets of patterns, and a pattern that is unlike any seen by the algorithm before is quite likely to be misinterpreted Human cognition / recognition cultural knowledge and experience Machine-learning methods generalize based on observed data

Recommendation Systems
Collaborative Filtering Weighted ranking of features The friend with good taste in recommends

You might also like