How supervision is arranged can affect the way that humans learn concepts. Yet very little is kno... more How supervision is arranged can affect the way that humans learn concepts. Yet very little is known about the role that supervision plays in nonhuman concept learning. Prior research in pigeon concept learning has commonly used differential response–reinforcer procedures (involving high-level supervision) to support reliable discrimination and generalization involving from 4 to 16 concurrently presented photographic categories. In the present project, we used contextual cueing, a nondifferential reinforcement procedure (involving low-level supervision), to investigate concept learning in pigeons. We found that pigeons were faster to peck a target stimulus when 8 members from each of 4 categories of black-and-white photographs— dogs, trees, shoes, and keys— correctly cued its location than when they did not. This faster detection of the target also generalized to 4 untrained members from each of the 4 photographic categories. Our results thus pass the prime behavioral tests of conceptualization and suggest that high-level supervision is unnecessary to support concept learning in pigeons.
We trained six pigeons in a stagewise Multiple Necessary Cues (MNC) go/no-go task to document the... more We trained six pigeons in a stagewise Multiple Necessary Cues (MNC) go/no-go task to document the dynamics of discrimination learning involving increasingly complex visual stimuli. The compound stimuli were composed from four dimensions, each of which could assume either of two extreme values or their intermediate value: Shape, Size, Line Orientation, and Brightness. Starting with a stimulus composed entirely from intermediate values, we replaced those values with each of the two extreme dimensional values in four successive stages, thereby increasing the stimulus set from 2 in Stage 1 to 16 in Stage 4. In each stage, only one combination of values signaled food (S +), whereas the remaining combinations did not (S − s). We calculated the rate of pecking during the first 15 s of each stimulus presentation and, in any given stage, training continued until the rate of responding to all of the S − s was less than 20% of the rate of responding to the S +. All pigeons successfully acquired the final discrimination, suggesting that they attended to all of the dimensions relevant for the discrimination. We also repli-cated the key results of prior MNC studies: (1) the number of extreme dimensional values in each stage was positively related to the amount of training required for pigeons to acquire the discrimination; (2) attentional tradeoffs were most often observed when three or four dimensions were being trained; and (3) throughout training, the number of dimensional values in the S − s that differed from the S + was positively related to their discriminability from S + .
We explored the relation between concept learning and the number of training exemplars, programme... more We explored the relation between concept learning and the number of training exemplars, programmed in a stepwise fashion. Eight pigeons (Columba livia) were trained via differential food reinforcement to peck 1 of 2 simultaneously displayed color images: benign or malignant human breast tissue samples. In each session, only 1 exemplar from each category was trained. If the learning criterion was achieved in 1 session, then those 2 exemplars were discarded and 2 new exemplars were trained in the next session. For the consistent group the reinforced category (benign or malignant stimuli for different birds) was maintained throughout all 24 training pairs, but for the inconsistent group the reinforced category was pseudorandomly changed from pair to pair. Pigeons in the consistent group evidenced striking gains in accuracy during the 1st few trials of training on new pairs (exceeding 90% correct over Pairs 10–12), whereas pigeons in the inconsistent group evidenced reliably weaker gains in initial categorization accuracy (rising above only 70% correct after comparable training). Stepwise training involving as few as a dozen exemplars per category was thus sufficient to support robust concept learning in pigeons.
Pathologists and radiologists spend years acquiring and refining their medically essential visual... more Pathologists and radiologists spend years acquiring and refining their medically essential visual skills, so it is of considerable interest to understand how this process actually unfolds and what image features and properties are critical for accurate diagnostic performance. Key insights into human behavioral tasks can often be obtained by using appropriate animal models. We report here that pigeons (Columba livia)—which share many visual system properties with humans—can serve as promising surrogate observers of medical images, a capability not previously documented. The birds proved to have a remarkable ability to distinguish benign from malignant human breast histopathology after training with differential food reinforcement; even more importantly, the pigeons were able to generalize what they had learned when confronted with novel image sets. The birds’ histological accuracy, like that of humans, was modestly affected by the presence or absence of color as well as by degrees of image compression, but these impacts could be ameliorated with further training. Turning to radiology, the birds proved to be similarly capable of detecting cancer-relevant microcalcifications on mammogram images. However, when given a different (and for humans quite difficult) task—namely, classification of suspicious mammographic densities (masses)—the pigeons proved to be capable only of image memorization and were unable to successfully generalize when shown novel examples. The birds’ successes and difficulties suggest that pigeons are well-suited to help us better understand human medical image perception, and may also prove useful in performance assessment and development of medical imaging hardware, image processing, and image analysis tools.
How supervision is arranged can affect the way that humans learn concepts. Yet very little is kno... more How supervision is arranged can affect the way that humans learn concepts. Yet very little is known about the role that supervision plays in nonhuman concept learning. Prior research in pigeon concept learning has commonly used differential response–reinforcer procedures (involving high-level supervision) to support reliable discrimination and generalization involving from 4 to 16 concurrently presented photographic categories. In the present project, we used contextual cueing, a nondifferential reinforcement procedure (involving low-level supervision), to investigate concept learning in pigeons. We found that pigeons were faster to peck a target stimulus when 8 members from each of 4 categories of black-and-white photographs— dogs, trees, shoes, and keys— correctly cued its location than when they did not. This faster detection of the target also generalized to 4 untrained members from each of the 4 photographic categories. Our results thus pass the prime behavioral tests of conceptualization and suggest that high-level supervision is unnecessary to support concept learning in pigeons.
We trained six pigeons in a stagewise Multiple Necessary Cues (MNC) go/no-go task to document the... more We trained six pigeons in a stagewise Multiple Necessary Cues (MNC) go/no-go task to document the dynamics of discrimination learning involving increasingly complex visual stimuli. The compound stimuli were composed from four dimensions, each of which could assume either of two extreme values or their intermediate value: Shape, Size, Line Orientation, and Brightness. Starting with a stimulus composed entirely from intermediate values, we replaced those values with each of the two extreme dimensional values in four successive stages, thereby increasing the stimulus set from 2 in Stage 1 to 16 in Stage 4. In each stage, only one combination of values signaled food (S +), whereas the remaining combinations did not (S − s). We calculated the rate of pecking during the first 15 s of each stimulus presentation and, in any given stage, training continued until the rate of responding to all of the S − s was less than 20% of the rate of responding to the S +. All pigeons successfully acquired the final discrimination, suggesting that they attended to all of the dimensions relevant for the discrimination. We also repli-cated the key results of prior MNC studies: (1) the number of extreme dimensional values in each stage was positively related to the amount of training required for pigeons to acquire the discrimination; (2) attentional tradeoffs were most often observed when three or four dimensions were being trained; and (3) throughout training, the number of dimensional values in the S − s that differed from the S + was positively related to their discriminability from S + .
We explored the relation between concept learning and the number of training exemplars, programme... more We explored the relation between concept learning and the number of training exemplars, programmed in a stepwise fashion. Eight pigeons (Columba livia) were trained via differential food reinforcement to peck 1 of 2 simultaneously displayed color images: benign or malignant human breast tissue samples. In each session, only 1 exemplar from each category was trained. If the learning criterion was achieved in 1 session, then those 2 exemplars were discarded and 2 new exemplars were trained in the next session. For the consistent group the reinforced category (benign or malignant stimuli for different birds) was maintained throughout all 24 training pairs, but for the inconsistent group the reinforced category was pseudorandomly changed from pair to pair. Pigeons in the consistent group evidenced striking gains in accuracy during the 1st few trials of training on new pairs (exceeding 90% correct over Pairs 10–12), whereas pigeons in the inconsistent group evidenced reliably weaker gains in initial categorization accuracy (rising above only 70% correct after comparable training). Stepwise training involving as few as a dozen exemplars per category was thus sufficient to support robust concept learning in pigeons.
Pathologists and radiologists spend years acquiring and refining their medically essential visual... more Pathologists and radiologists spend years acquiring and refining their medically essential visual skills, so it is of considerable interest to understand how this process actually unfolds and what image features and properties are critical for accurate diagnostic performance. Key insights into human behavioral tasks can often be obtained by using appropriate animal models. We report here that pigeons (Columba livia)—which share many visual system properties with humans—can serve as promising surrogate observers of medical images, a capability not previously documented. The birds proved to have a remarkable ability to distinguish benign from malignant human breast histopathology after training with differential food reinforcement; even more importantly, the pigeons were able to generalize what they had learned when confronted with novel image sets. The birds’ histological accuracy, like that of humans, was modestly affected by the presence or absence of color as well as by degrees of image compression, but these impacts could be ameliorated with further training. Turning to radiology, the birds proved to be similarly capable of detecting cancer-relevant microcalcifications on mammogram images. However, when given a different (and for humans quite difficult) task—namely, classification of suspicious mammographic densities (masses)—the pigeons proved to be capable only of image memorization and were unable to successfully generalize when shown novel examples. The birds’ successes and difficulties suggest that pigeons are well-suited to help us better understand human medical image perception, and may also prove useful in performance assessment and development of medical imaging hardware, image processing, and image analysis tools.
Uploads
Books by Ed Wasserman
Papers by Ed Wasserman