Partially Verified65% confidenceFactExact time
The goal of supervised learning is to find a function f that, for a given input x, can predict the corresponding output y.
3
Sources
65%
Confidence
Long-term
Relevance
8/2/2026
First Seen
Sources
Related Entities
Related Claims
Unverified监督学习的目标是学习输入特征X到输出标签Y的映射函数，分为分类（离散标签）和回归（连续标签）两大子任务73% similarUnverified神经网络训练的目标是根据已知的一组x和y猜出所有的w和b参数71% similarUnverified现代机器学习模型本质上是复杂的数学函数 f(x)=y，其中 x 是输入特征向量，y 是预测输出，f 是从训练数据中学习到的映射关系66% similarUnverifiedFinding the function mapping in supervised learning is essentially an optimization problem that adjusts model parameters by minimizing the loss function (difference between predicted and true values).66% similarUnverified预训练的核心目标是让模型具备预测下一个词（Predict Next Token）的能力，基于Transformer架构中的因果注意力机制62% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/679931API
curl https://kongchang.com/api/v1/knowledge/claims/679931MCP
get_claim(id=679931)