求问狗家两道面试题目

狗家的ml design两道题目 请问一下大家有什么思路？先谢过了
1. we have annual training data of temperature and number of death due to a disease. How do we make monthly predictions?
2. Say you have a 1000 labeled data points and every day you get a large batch of data that are not labeled. Getting the labels are expensive. How do you find the top examples to get a label and feed that into the subsequent training loop?

第一题简单的supervised learning吧。
第二题是semisupervised learning，可以KNN先label然后再train？

微信用户_k3yy2 发表于 2023-11-14 14:12
第一题简单的supervised learning吧。
第二题是semisupervised learning，可以KNN先label然后再train？

第一题问题是要monthly prediction哦 有思路吗？
第二题就是semisupervised learning是正解

年度数据有温度，就对温度做个回归之类的，预测月的话，再找一下各个月的温度？