Learning by observing others' experiences is a hallmark of human intelligence. While the neurocomputational mechanisms underlying observational learning are well understood, less is known about whom people prefer to learn from in the context of observational learning. One hypothesis posits that learners prefer individuals who exhibit a high degree of decision noise, 'free riding' on the costly exploration of others. An alternative hypothesis is that learners prefer individuals with low decision noise, as lower decision noise is often associated with better performance. In a preregistered experiment, we found that most participants preferred to learn from low-noise (high-performing) individuals. Furthermore, exploratory analyses revealed that participants who preferred low-noise individuals tended to rely on imitation of others' actions. These findings offer a potential computational account of how learning styles are related to partner selection in social learning.

In display advertising, predicting the conversion rate (CVR), meaning the probability that a user takes a predefined action on an advertiser's website, is a fundamental task for estimating the value of displaying an advertisement to a user. There are two main challenges in CVR prediction due to delayed feedback. First, some positive labels are not correctly observed in training data because some conversions do not occur immediately after a click. Second, delay mechanisms are not uniform among instances, meaning some positive feedback are much more frequently observed than others. It is widely acknowledged that these problems lead to severe bias in CVR prediction. To overcome these challenges, we propose two unbiased estimators: one for CVR prediction and the other for bias estimation. Subsequently, we propose a dual learning algorithm in which a CVR predictor and a bias estimator are trained in alternating fashion using only observable conversions. The proposed algorithm is the first of its kind to address the two major challenges in a theoretically sophisticated manner. Empirical evaluations using synthetic datasets demonstrate the practical value of the proposed approach.

In display advertising, predicting the conversion rate, that is, the probability that a user takes a predefined action on an advertiser’s website, such as purchasing goods is fundamental in estimating the value of displaying the advertisement. However, there is a relatively long time delay between a click and its resultant conversion. Because of the delayed feedback, some positive instances at the training period are labeled as negative because some conversions have not yet occurred when training data are gathered. As a result, the conditional label distributions differ between the training data and the production environment. This situation is referred to as a feedback shift. We address this problem by using an importance weight approach typically used for covariate shift correction. We prove its consistency for the feedback shift. Results in both offline and online experiments show that our proposed method outperforms the existing method.