A large healthcare system wants to automatically generate patient education summaries based on long clinical notes. They have millions of unlabeled documents, but no “correct summaries” to train on — so they choose a model that can learn language patterns by predicting the next word and then generate new text.

What type of learning approach are they using?

Question

A large healthcare system wants to automatically generate patient education summaries based on long clinical notes. They have millions of unlabeled documents, but no “correct summaries” to train on — so they choose a model that can learn language patterns by predicting the next word and then generate new text.

What type of learning approach are they using?

BlackTom AI · Accepted Answer

The scenario describes generating patient education summaries from long clinical notes when there are millions of unlabeled documents, and the model learns language patterns by predicting the next word to generate new text. This setup points to a specific learning paradigm in machine learning.
Option 1: Supervised Learning. In supervised learning, the model learns from input-output pairs with explicit labels (e.g., input features paired with target summaries). Here there are no labeled summaries to train on, so this option does not fit the described method.
Option 2: Self-Supervised Learning. This approach uses the data itself to create learning signals, such as predicting the next word or missing pieces, without requiring human-provided labels. The description matches next-word prediction on unlabeled clinical notes, which is a classic self-supervised objective in language modeling.
Option 3: Unsupervised Learning. While related, unsupervised learning typically involves discovering structure or representations without any predictive target (e.g., clustering or dimensionality reduction). Predicting the next word is a predictive task, not just unsupervised structure discovery, so this option is incomplete or misleading for this specific context.
Option 4: Reinforcement Learning. Reinforcement learning relies on interactions with an environment to maximize cumulative rewards, which is not described here. The task is language modeling from text data, not learning via trial-and-error with rewards.

类似问题

Self-supervised learning differs from pure supervised learning in that the labels used by the learning process are contained in the data itself.

Match each method with its key characteristic: 1: MoCo 2: SimCLR 3: BYOL 4: SwAV

Name the Supreme Court decision that removed financial constraints on PACs.

Political scientists Joshua Kalla and David Broockman find that senior policy makers made themselves available to _____ between three and four times more often than they did to constituents.

Name the political scientist who argues that parties are designed to serve the needs of ambitious politicians (namely candidates and officeholders).

Name the political scientist who said this famous quote about the problems of pluralism: “The flaw in the pluralist heaven is that the heavenly choir sings with a strong upper-class accent.”

What is the name of the third party candidate in 2000 who was not allowed to participate in the presidential debates? Hint: We watched a video clip with him in class.

更多留学生实用工具

智能学习助手

风格化写作助手

论文查重助手

文献引用助手

课堂转译助手

课堂笔记助手

Quiz搜索助手

学校历年真题

智能刷题助手

智能匹配练习

希望你的学习变得更简单