Artificial Intelligence / Machine Learning Methods

Also: @Robots. @Prolog. @Expert Systems. Note: We don't actually have Artificial Intelligence (yet). We have Machines that Learn.

Methods

• Linear Regression "Draw a line through it" Remember y = mx + b?
• Logistic Regression Replace the line with a quick move to a limit,
  Classifier (sample) Divide data points into classes
• Support Vector Machine (SVM) Draw the best possible line between classes
• Clustering (K-Means) Figure out what the classes are
• Genetic Algorithms
• Anomaly Detection Detect outliers
• Prediction / Recommendation
• Localization (computing position over time from sensor readings)
• A * Search
• Large Datasets
• Bayesian Classification Find the probable center of data clusters
• Eigenvector / Eigenvalue
• Neural Networks
  • Convolutional Neural Networks
  • CMAC:Cerebellar Model Articulation Controller
  • Neural Networks vs Genetic Algorithms
• Troubleshooting: Bias vs Variance, Under/Over Fitting, Stastical Issues (Skewed data)
  • Regularization,
  • Principal Component Analysis (PCA)
  • Feature Normalization
• Image Recognition
• Natural Language processing.
• STRIPS: "STanford Research Institute Problem Solver"

Research:

• Lyapunov functions. Because we always want the derivative of our cost function to smoothly decend to a singal minimum.

Also:

• July 2003 SXList.com newsletter - SX Learning Robotics
• The ability of the human mind to recognize mis-spelled words is a good example of the current division between software and wetware.
• Why your job is NOT comming back

See also:

• http://aima.cs.berkeley.edu/ THE textbook.
• https://youtu.be/v9M2Ho9I9Qo Excellent AI video. How to learn without examples.
• https://medium.com/explore-artificial-intelligence/tagged/machine-learning Good series of articles.
• https://github.com/oxford-cs-deepnlp-2017/lectures Oxford Deep Learning 2017
• https://docs.google.com/spreadsheets/d/1AQvZ7-Kg0lSZtG1wlgbIsrm90HaTZrJGQMz-uKRRlFw/edit List of datasets. Sample data from many different areas to use in training AI systems.
• http://course.fast.ai/ Free Deep Learning course from University of San Francisco
• https://github.com/libelo/py-ng-ml Repository of Python versions of common Machine Learning methods from a popular online course.
• https://arxiv.org/list/cs.AI/recent Fully accesable research papers on AI.
• http://spark.apache.org/ Apache Sparkâ„¢ is a free, open source, fast and general engine for large-scale data processing. Spark runs on Hadoop, Mesos, standalone, or in the cloud. It can access diverse data sources including HDFS, Cassandra, HBase, and S3. Write applications quickly in Java, Scala, Python, R. Run programs up to 100x faster than Hadoop MapReduce in memory, or 10x faster on disk.
• http://www.elcct.com/installing-hadoop-2-3-0-on-ubuntu-13-10/ The Apache Hadoop software library is a framework that allows for the distributed processing of large data sets across clusters of computers using simple programming models. It is designed to scale up from single servers to thousands of machines, each offering local computation and storage. Rather than rely on hardware to deliver high-availability, the library itself is designed to detect and handle failures at the application layer, so delivering a highly-available service on top of a cluster of computers, each of which may be prone to failures.
• https://www.youtube.com/watch?v=qDbpYUbf3e0 very high level overview of Machine Learning.
• https://www.youtube.com/watch?v=E8SEVsxV070 the XOR function makes fitting a line to the data impossible. Many times, in the real world, conditions were extreems are bad, or good and the middle condition is the opposite look like the XOR problem. Sometimes you can learn the wrong answer faster than the right answer.
• https://www.youtube.com/watch?v=l42lr8AlrHk Why it's good to have extra layers in your learning system: Deep Learning.
• http://www.scaruffi.com/mind/ai.html History of AI.
• https://www.youtube.com/watch?v=BCBZPtZCI7w Facial recognition with Machine Learning.
• http://seit.unsw.adfa.edu.au/staff/sites/dcornforth/CMAC.html CMAC for Classification
• http://numenta.org/ NuPIC is an open source project based on a theory of neocortex called Hierarchical Temporal Memory (HTM)
• https://archive.org/stream/byte-magazine-1979-07-rescan/1979_07_BYTE_04-07_Automating_Eclipses Byte Magazine July 1979. "A Model of the Brain for Robot Control. Part 2" By James Albus, inventor of the CMAC. Gives an overview of the idea and it's origins.
• https://archive.org/stream/byte-magazine-1979-08-rescan/1979_08_BYTE_04-08_LISP Byte Magazine August 1979. "A Model of the Brain for Robot Control. Part 3" By James Albus, inventor of the CMAC. Tests the concept and expands on it's uses.
• http://www.cs.cmu.edu/afs/cs/project/ai-repository/ai/areas/neural/systems/cmac/ Simple version of the CMAC C code. Easy to follow.
• http://www.ece.unh.edu/robots/cmac.htm C code for the CMAC neural net alternative by James Albus. cache 2012

• http://www.bvandam.net/
• http://home.infi.net/~wtnewton/otherwld/selfwire.html simulating a nervous network on a processor A Self-Wiring Array of "Bicores" for Robotic Control (cached 20010404103345 ) 
• http://www.merlotti.com/EngHome/Computing/AntsSim/ants.htm Simulation of Ants
• http://www.amristar.com.au/~hutch/hex/ A very good faker. Winner of the 1996 Turing test.
• https://www.youtube.com/watch?v=gfY2LfRfE1E Overview of Machine learning with some technical details from SIGGRAPH 2020.