5 Tribes of Machine Learning – BMC Software | Blogs
In Pedro Domingos’ book, The Master Algorithm: How The Quest for the Ultimate Learning Machine Will Remake Our World, he categorizes the types of machine learning algorithms into five classes, which he calls the tribes of machine learning. Each group supports a set of principles, and, from them, stem different machine learning models.
The "master algorithm" is the algorithm that they will converge on, the algorithm that will make
Artificial general intelligence a.k.a. strong AI, full AI, human-level AI, and general intelligent action.
Symbolists - explicit inference rules, often top-down. The dominant paradigm of AI from its 1950's beginnings to the 1980's. No longer as dominant because of the success of bottom-up models in many problems.
Inference rules these very simplified ones for traveling and traffic lights:
- If red, then stop
- If yellow, then if relatively far, then stop, else go
- If green, then go
Notice the additional test for a yellow light. That threshold may be determined in bottom-up fashion, creating a hybrid approach.
-- Decision trees, Random decision forests, Production rule systems, Inductive logic programming
All the rest are bottom-up, fas far as I can tell.
Connectionists (Neuroscience) - modeled on how nervous systems work, including brains. The connections are between parts that do simple things. A big problem is the difficulty of interpreting its parameter values.
-- Artificial neural nets, Reinforcement learning, Deep learning
Bayesians (Statisticians) - does probabilistic modeling, though often with other types of models: symbolic, connectionist.
-- Hidden Markov chains, Graphical models, Causal inference
Evolutionaries (Biologists) - "Looking at machine learning from a biologist stance, they are concerned with the evolution of an AI. They’re curious with how it grows, mutates…with how it becomes." and "Evolutionaries use genetic algorithms, and evolutionary programming. A common application of Evolutionary AI is on learning tasks."
Analogizers (Psychologists) - "Analogizers are usually the storytellers. They can create classes of entities. If an input, old or new, is identified as part of one of those classes, then Analogizers believe they can predict the outcome of the input as being like the outcome of that class." They like to use clustering algorithms.
Each approach has its usefulness. Combinations of these Machine Learning methods might be the answer to an AGI. Self-driving cars might learn to drive safely on the roads through an Evolutionary method, but they will use Connectionist methods to give the cars’ sensors sight. The car might get a huge boost in its user-driver interactions by utilizing the Analogizer’s way of classifying its drivers into types, Aggressive, Defensive, or Passive. Finally, there are rules to the road such as stopping at stop signs and staying between the lines, where methods from the traditional Symbolist tribe prove beneficial.
