Mertens, G., Lodder, P., Smeets, T., & Duijndam, S. (2021). Fear of COVID-19 predicts vaccination willingness 14 months later. PsyArXiv. https://doi.org/10.31234/osf.io/rt7u4

SARS-CoV-2 variants of concern and variants under investigation. (2021). 45.

Trevor Bedford. (2021, December 11). We find that logistic growth of Omicron sequence fraction looks similar between the UK, the US and Germany with roughly 1% of sequenced cases in all three countries being Omicron on Dec 1. 3/10 https://t.co/De0t2xreU9 [Tweet]. @trvrb. https://twitter.com/trvrb/status/1469784757261127685

Liu, Y., Finch, B. K., Brenneke, S. G., Thomas, K., & Le, P. D. (2020). Perceived Discrimination and Mental Distress Amid the COVID-19 Pandemic: Evidence From the Understanding America Study. American Journal of Preventive Medicine, 0(0). https://doi.org/10.1016/j.amepre.2020.06.007

Ip, A., Ahn, J., Zhou, Y., Goy, A. H., Hansen, E., Pecora, A. L., Sinclaire, B. A., Bednarz, U., Marafelias, M., Mathura, S., Sawczuk, I. S., Underwood, J. P., Walker, D. M., Prasad, R., Sweeney, R. L., Ponce, M. G., LaCapra, S., Cunningham, F. J., Calise, A. G., … Goldberg, S. L. (2020). Hydroxychloroquine in the treatment of outpatients with mildly symptomatic COVID-19: A multi-center observational study. MedRxiv, 2020.08.20.20178772. https://doi.org/10.1101/2020.08.20.20178772

Sun, K., Wang, W., Gao, L., Wang, Y., Luo, K., Ren, L., Zhan, Z., Chen, X., Zhao, S., Huang, Y., Sun, Q., Liu, Z., Litvinova, M., Vespignani, A., Ajelli, M., Viboud, C., & Yu, H. (2020). Transmission heterogeneities, kinetics, and controllability of SARS-CoV-2. MedRxiv, 2020.08.09.20171132. https://doi.org/10.1101/2020.08.09.20171132

Gaiha, S. M., Cheng, J., & Halpern-Felsher, B. (2020). Association Between Youth Smoking, Electronic Cigarette Use, and Coronavirus Disease 2019. Journal of Adolescent Health, 0(0). https://doi.org/10.1016/j.jadohealth.2020.07.002

Ahn, M. H., Shin, Y. W., Kim, J. H., Kim, H. J., Lee, K.-U., & Chung, S. (2020). High Work-related Stress and Anxiety Response to COVID-19 among Healthcare Workers in South Korea: SAVE study [Preprint]. PsyArXiv. https://doi.org/10.31234/osf.io/9nxth

Samuel, J.; Ali, G.G.M.N.; Rahman, M.M.; Esawi, E.; Samuel, Y. COVID-19 Public Sentiment Insights and Machine Learning for Tweets Classification. Preprints 2020, 2020050015 (doi: 10.20944/preprints202005.0015.v1)

Roland, L. T., Gurrola, J. G., Loftus, P. A., Cheung, S. W., & Chang, J. L. (2020). Smell and taste symptom‐based predictive model for COVID‐19 diagnosis. International Forum of Allergy & Rhinology, alr.22602. https://doi.org/10.1002/alr.22602

生成模型 vs. 判别模型

总体来看，如果样本足够多，判别模型的正确率高于生成模型的正确率。

生成模型和判别模型最大的区别在于，生成模型预先假设了很多东西，比如预先假设数据来自高斯，伯努利，符合朴素贝叶斯等等，相当于预先假设了 Hypothesis 函数集，只有在此基础上才有可能求出这个概率分布的参数。

生成模型，进行了大量脑补。脑补听起来并不是一件好事，但是当你的数据量太小的时候，则必须要求你的模型具备一定的脑补能力。

判别模型非常依赖样本，他就是很传统，死板，而生成模型比较有想象力，可以“想象”出不存在于当前样本集中的样本，所以他不那么依赖样本。

关于 想象出不能存在于当前样本集的样本 ，见本课程 40:00 老师举例。

生成模型在如下情形比判别模型好：

1. 数据量较小时。
2. 数据是noisy，标签存在noisy。
3. 先验概率和类别相关的概率可以统计自不同的来源。

释疑第三条优点：老师举例，在语音辨识问题中，语音辨识部分虽然是 DNN --- 一个判别模型，但其整体确实一个生成模型，DNN 只是其中一块而已。为什么会这样呢？因为你还是要去算一个先验概率 --- 某一句话被说出来的概率，而获得这个概率并不需要样本一定是声音，只要去网络上爬很多文字对话，就可以估算出这个概率。只有 类别相关的概率 才需要声音和文字pair，才需要判别模型 --- DNN 出马。

Interpreting the reference group in dummy coding.

the development and success of capitalism benefits urbanization by contributing resources over time that spur its growth.