统计机器学习笔记.zip

1. Frequentist (频率派): The Frequentist approach views the model parameters as unknown (constant) and estimate them by matching the model to the training data. Using an appropriate metric. 频率派观点认为,模型参数是未知的(常数),用恰当的准则,使得模型与训练数据相匹配来估计它们 (参数)。 2. Bayesian (⻉叶斯派): ⻉叶斯派观点认为,模型参数是随机变量,通过给予参数以先验分布(通常是来自前人经验)