Artificial inteligence without activation function #19236
Replies: 3 comments 5 replies
-
This seems a very general question. AI is not well-defined at all. Is a decision tree AI? In which case the answer to your question is yes. My feeling as someone who is hoping that Github Discussions can be useful for scikit-learn, there is a threshold where we should feel free say "this is not the best place to ask this kind of questions" meaning "you probably are not going to get a great answer here" and "here is a better place to ask this kind of questions". The threshold is not well-defined and I don't really know where this threshold is but hopefully with time we can figure it out! |
Beta Was this translation helpful? Give feedback.
-
going out of boundary, will lead to some astonishing discovery, i have new approach to perform AI, using a new tool to collect patterns from the training data set in order to use theme to validate the test data set, fully transparent, no black box |
Beta Was this translation helpful? Give feedback.
-
Activation functions, weight and biases are concepts typically used to describe neural network architectures used in Deep Learning. Machine Learning in general deals with a larger class of predictive models that are not limited to neural network models. For instance a very popular models are decision tree-based models such as Gradient Boosting Trees which are very different from neural networks. Furthermore, Artificial Intelligence is a wide research field that encompasses many computer science domains, not limited to machine learning in general and deep learning in particular. Such sub-fields of Artificial Intelligence could include Operations Research, typically to solve complex scheduling or planning problems via a mix of discrete and continuous optimization and search algorithms and symbolic reasoning (deductive, abductive or inductive) for instance to design automated theorem provers and formal model checking. To summarize: neural networks (deep learning) is a sub-class of machine learning models and machine learning models are just one way (although a very popular one) to design software systems which exhibit some behaviors that operators could consider somewhat artificially intelligent. |
Beta Was this translation helpful? Give feedback.
-
This is my question.
Is it possible to perform AI, without a neural network, activation fonction weight and bias ?
Beta Was this translation helpful? Give feedback.
All reactions