Search Results
7/11/2025, 6:46:40 PM
>>510103833
It's as hard as downloading chrome from source and trying to make heads or tail of the v8 engine around video, so it's obfuscated as people are protecting their propriatary information, but broad strokes are harder to hide. It's a machine learning algorithm applying machine learning algorithms to problems. The algorithms submitted have to obey a certain interface, that is they have to have a data load step, a data organization step, then the training/convergence step and produce a final model, then the model is tested alone, and "models are hybrid with models" grouped by machine learning algorithm, and evaluated for correctness. So there are tens of thousands of machine learning algorithms in the hopper to be chosen. The head algorithm of ChatGPT basically uses machine learning to replace the machine learning developer's job: Data Gathering, Feature Extraction, Data Cleaning, Data Labeling, Data Representation, Apply transforms for Missing Data, Automatic feature refinement, Automatic feature creation, Automatic feature rating, Data Visualization, Automatic feature mating, K-Means Clustering analysis, Data Reduction, Principal Components Analysis, collapsing statistically redundant data, Apply a dozen efficient brute force enumerated known truth algorithms like sorting or filtering or translating images to text or text to images, or both, to image and back again to figure out what monkeymind says when this word is piped through image search., Then application of machine learning classifiers to get from where you are to where you need to be Decision tree, neural networks, Support vector machines, Bagging/Boosting Qlearning, thousands of variations that are only understood by the people who made them. Discrete, Continuous and Binary classifiers.
It's as hard as downloading chrome from source and trying to make heads or tail of the v8 engine around video, so it's obfuscated as people are protecting their propriatary information, but broad strokes are harder to hide. It's a machine learning algorithm applying machine learning algorithms to problems. The algorithms submitted have to obey a certain interface, that is they have to have a data load step, a data organization step, then the training/convergence step and produce a final model, then the model is tested alone, and "models are hybrid with models" grouped by machine learning algorithm, and evaluated for correctness. So there are tens of thousands of machine learning algorithms in the hopper to be chosen. The head algorithm of ChatGPT basically uses machine learning to replace the machine learning developer's job: Data Gathering, Feature Extraction, Data Cleaning, Data Labeling, Data Representation, Apply transforms for Missing Data, Automatic feature refinement, Automatic feature creation, Automatic feature rating, Data Visualization, Automatic feature mating, K-Means Clustering analysis, Data Reduction, Principal Components Analysis, collapsing statistically redundant data, Apply a dozen efficient brute force enumerated known truth algorithms like sorting or filtering or translating images to text or text to images, or both, to image and back again to figure out what monkeymind says when this word is piped through image search., Then application of machine learning classifiers to get from where you are to where you need to be Decision tree, neural networks, Support vector machines, Bagging/Boosting Qlearning, thousands of variations that are only understood by the people who made them. Discrete, Continuous and Binary classifiers.
Page 1