**5.6. Prediction of the next question**

**5. Challenges of using AML for human categorization**

Most of the empirical studies, which addressed human HAML, focus on learning two-class problems on a one-dimensional input space [15, 29], but there are obstacles to generalize the model to multiclass classification problems [19]. The same is true for a single-modality (visual) object recognition. It is important to investigate the generalization of human AML to multidimensional objects (stimuli), such as auditory stimuli, high-dimensional stimuli, real-world

Experimental studies on the human AML show that it is sensitive to noise and humans are not as good as machines in selecting queries from an unlabeled dataset of artificial visual stimuli [21]. Thus, human AML performance declines with higher noise levels. Humans perform relatively well in at least some noise settings, suggesting that they took the experiment

Zhu et al. [22] showed that humans are sensitive to the distributional structure of the subsequent unlabeled experience. Gibson et al. [19] investigated the effects of the distributions of unlabeled instances (stimuli) to human learner in two experiments, and they also investigated the effect of the order of the unlabeled items that participants encountered in an experiment. They concluded that human categorization is sensitive to both the distribution and ordering

The small number of participants limits the generalization of the findings to other humans. In many studies that investigated human AML, small group of people participated in the experiments. In addition, a small number of objects used in the investigations lead to a similar limitation, because a limited number of teaching and test instances reduce the reliability of

People are sensitive to the value of both labeled and unlabeled stimuli, and this depends on the structure of the concept being learned [24]. Markant and Gureckis [31] showed that the effectiveness of AML might interact with the particular structure of the target categories. Two types of category structures were used in that study: rule-based (RB), in which the decision rule is defined as a criterion along a single dimension, and information integration, in which

**5.1. High dimensionality**

24 Active Learning - Beyond the Future

**5.2. Sensitivity to noise**

of unlabeled instances [19].

**5.5. Category or concept structure**

seriously [30].

the results.

stimuli, other demographic groups, etc.

**5.3. Distribution and ordering of unlabeled data**

**5.4. Small number of participants and objects**

the decision rule is a function of at least two dimensions.

In human learning, people often learn by asking rich and interesting questions, which more directly target the concepts in a learning task. For example, a child might ask "Do all dogs have long tails?" or "What is the difference between cats and dogs?" [32]. The main challenge for AML method is to predict which question a human will ask from the given context. A number of recent studies have discussed this challenge [33–35]. Rothe et al. in [33] proposed a model that predicts what questions human learners will ask and can creatively generate novel questions that did not exist in the training data. Their work in [34] showed that human can accurately evaluate question quality by using the Bayesian ideal observers. In the most recent review [35], authors highlight and discuss nine challenges about the psychology of human inquiry.
