• View
• Edit
• History
• Print

Artificial Intelligence matrix

see also DATAmatrix

Principle

Based on this sumarized layout artificial intelligence is a key component of Seedea. In order to maximize the cognitive ROI the best framework has to be chosen. The objective of this page is to be able to select it more easily.

Categories

• AI integrated unifying frameworks (OpenCog, ...)
• AI libraries (http://www.cliki.net/AI, ...)
• Semantic engines (Twine, ...)
• directly usable for end-user : out of scope of this page (but can be interesting to find the components they are using)

Result

To add

• personal selections I did during 1st sem 2008
• http://www.cliki.net/AI
• API of the new popular semantics engines (see AlternativeSearchEngines list/pop web3.0 coverage)
• scipy, ffnet, networkx
• freshmeat repository
• RSS feeds on news of the domain (new framework should be automatically added to the matrix)
• http://en.wikipedia.org/wiki/Special:Search/Comparison_of
• WEKA : Waikato Environment for Knowledge Analysis, suggested by JaredWigmore on #opencog
  • standard machine learning algorithms (toolkit-style)
    • classification, clustering, association rule mining, attribute selection, ...
• Category:Cognitive architecture on Wikipedia : cognitive architectures, models of human cognition.
• Category:AGI Projects on Artificial General Intelligence Research Institute
• Improving Software Productivity and Quality via Mining Program Source Code from North Carolina State University
  • based on the Lisp principle that code is data and data is potential code
• MLOSS : machine learning open source software with discussion at NIPS ´08 Workshop: Machine Learning Open Source Software MLOSS´08 - Whistler
• PowerLoom language and environment for constructing intelligent, knowledge-based applications
• Linas' Artificial General Intelligence list
  • Computational Linguistics and Natural Language Processing
• Sentiment Analysis and Language Processing Tools
• Cognitive Code Corporation and the SILVIA platform
• Potential results from the AGI Roadmap created by Scott Livingston and Itamar Arel in February 2009 and presented during AGI-09
• GNU/Linux AI & Alife HOWTO by John Eikenberry
• comparison of cognitive architectures according to Wikipedia
• SEMPRE: Semantically Aware Profiling for Recommenders by OFAI Language Technology Group
  • discovered while checking OFAI members whose director edited the Programming For Peace book
• Marcus' (Hutter) Artificial Intelligence Page
  • Universal Algorithmic Intelligence: A mathematical top->down approach (aka the AIXI model)
• AGIRI.org Artificial General Intelligence Research Institute
• Analytics1305 running over AmazonEC2
  • All Neighbors beta
  • Kernel Density Estimation
  • K-means
  • Support Vector Machines
• LaTeX Search Mathematical Equations in Scientific Publications, Springer
• MLcomp objectively comparing machine learning programs across various datasets for multiple problem domains.
• Kaggle supports two competition types: 'predicting the future' and 'predicting the past'.

Alternatives

• http://monicasmind.com/?cat=10 model free
• http://www.nature.com/nature/journal/v455/n7209/edsumm/e080904-01.html
• http://research.microsoft.com/en-us/collaboration/fourthparadigm/
• http://en.wikipedia.org/wiki/Big_data
• basically what Google does (1st data then somehow fit whatever model is possible on it, cf Peter Norvig)

To add to BackEnd with also heuristics and metaheuristics

To do

• Consider tablefy to organize the comparison or Google Squared
• subscribe to RSS updates (blog/code repository/wiki page) of currently active "Science Discovery" solutions (used in article like in A*****y university for "the scientific robot" or in AAAI article on the topic)
  • define clearly if they support the CAEE perspective
  • eventually make it a dedicated page
• http://www.AGI-Wiki.org/ formerly http://www.benetou.fr/agi/