Machine Learning Introduction _ Machine Learning

I. Working methods of machine learning

① Select data: Divide your data into three groups: training data, validating data, and testing data

② model data: Using training data to build models using related features

③ validation Model: Using your validation data to access your model

④ Test Model: Use your test data to check the performance of the validated model

⑤ Use model: Use fully trained models to make predictions on new data

⑥ Tuning Model: Improves performance of algorithms using more data, different features, or adjusted parameters

Two. Five major schools

①: Using symbols, rules, and logic to characterize knowledge and logical reasoning, the favorite algorithm is: Rules and decision tree

② Bayes: Get the possibility of occurrence to do probabilistic inference, favorite algorithm is: Naive Bayesian or Markov

③: Using probabilistic matrices and weighted neurons to dynamically identify and induce patterns, the favorite algorithm is: Neural networks

④ Evolution: Generate changes, then get the best of them for a particular goal, the favorite algorithm is: Genetic algorithm

⑤analogizer: Optimize functions according to constraints (go as high as possible, but don't leave the road at the same time), the favorite algorithm is: Support vector machine