In Search of Intelligence The Triumph of AI Devesh Rajadhyax CEO, Cere Labs
In Search of Intelligence
The Triumph of AI
Devesh Rajadhyax
CEO, Cere Labs
The story of AI started in 1956
...and it has seen many ups and downs
The initial exuberance …. late 50’s and 60’s
The confidence...
Achievements - Theory Proving program, beating humans at Checkers and so on.
...and then the depression of 70’s and late 80’s
The AI Winter
The quiet work during the winter in universities and some companies
Learning from our brain- 90’s the decade of brain
Readjustment of the AI vision
...from controlling the world to providing back end capabilities
The 2000’s : the making of an industry
Rise of the AI Startups
About Cere Labs
www.aimumbai.org [email protected]
AIM - a community for AI enthusiasts in Mumbai
Prediction/anomaly demo
Next…
The Brain &
Anomaly Detection Devesh
Siddhesh Neelam Harsh
➢ Why study the brain➢ Hierarchical Temporal Memory (HTM) ➢ Prediction and Anomaly Detection➢ Get started with NuPIC➢ Demo application
Agenda
Why is it important to study the brain?
“There is no scientific study more vital to man than the study of his own brain. Our entire view of the universe depends on it.”― Francis Crick
The neuron is the structural and functional unit of the nervous system - Santiago Ramón y Cajal
Layers of Neocortex
retina
cochlear
somatic
The neocortex learns a model from fast changing sensory data
The model generates● predictions● anomalies● actions
light
sound
touch
“The neocortex is uniform” - Vernon Benjamin Mountcastle
“It is the ability to make predictions about the future that is the crux of intelligence.”― Jeff Hawkins, On Intelligence
Vision:1. Discover operating principles of neocortex.2. Create technology for machine intelligence
based on neocortex principles.
HTM
Hierarchical Temporal Memory (HTM)
HTM is a framework that exploits three principles of brain
● Grouping patterns that occur together frequently.● Complex concepts from simple concepts.● Common cortical algorithm
Learning
Stages of HTM
Inference
Applications of HTM
Prediction Anomaly Detection
What is an anomaly?Something that deviates from what is standard, normal, or predicted.
Why anomaly detection?
Implementation of HTM using NuPIC
About NuPIC
● Numenta Platform for Intelligent Computing● Python based● Open Source
Applications based on HTM in NuPIC
CPU Usage Prediction/Anomaly System
CPU PREDICTION & ANOMALY DETECTION SYSTEM
● It is an application that attempts to predict CPU utilisation on the local system.
● It uses the NuPIC Online Prediction Framework (OPF), to derive predictions in real-time.
System Components:
Actual and Predicted CPUUsage
CPU.PY
NUPIC OPF
CPU USAGE
Model_param.py
System Requirements:
● Windows/Ubuntu/MAC
● Python development environment
● Nupic Libraries
● Matplotlib Libraries(for plotting the statistics on a graph)
Demonstration
Thank You!
References:
● http://numenta.com/learn/● Chandola V., Banerjee, A. and Kumar A., 2009. "Anomaly detection: A
survey." ACM computing surveys (CSUR), 41(3), p.15.● https://github.com/numenta/nupic/tree/master/examples/opf/clients/cpu