Duel-based Deep Learning system for solving IQ tests

Paulina Tomaszewska, Adam Żychowski, Jacek Mańdziuk
Proceedings of The 25th International Conference on Artificial Intelligence and Statistics, PMLR 151:10483-10492, 2022.

Abstract

One of the relevant aspects of Artificial General Intelligence is the ability of machines to demonstrate abstract reasoning skills, for instance, through solving (human) IQ tests. This work presents a new approach to machine IQ tests solving formulated as Raven’s Progressive Matrices (RPMs), called Duel-IQ. The proposed solution incorporates the concept of a tournament in which the best answer is chosen based on a set of duels between candidate RPM answers. The three relevant aspects are: (1) low computational and design complexity, (2) proposition of two schemes of pairing up candidate answers for the duels and (3) evaluation of the system on a dataset of shapes other than those used for training. Depending on a particular variant, the system reaches up to $82.8%$ accuracy on average in RPM tasks with 5 candidate answers and is on par with human performance and superior to other literature approaches of comparable complexity when training and test sets are from the same distribution.

Cite this Paper

BibTeX
@InProceedings{pmlr-v151-tomaszewska22a, title = { Duel-based Deep Learning system for solving IQ tests }, author = {Tomaszewska, Paulina and \.Zychowski, Adam and Ma\'ndziuk, Jacek}, booktitle = {Proceedings of The 25th International Conference on Artificial Intelligence and Statistics}, pages = {10483--10492}, year = {2022}, editor = {Camps-Valls, Gustau and Ruiz, Francisco J. R. and Valera, Isabel}, volume = {151}, series = {Proceedings of Machine Learning Research}, month = {28--30 Mar}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v151/tomaszewska22a/tomaszewska22a.pdf}, url = {https://proceedings.mlr.press/v151/tomaszewska22a.html}, abstract = { One of the relevant aspects of Artificial General Intelligence is the ability of machines to demonstrate abstract reasoning skills, for instance, through solving (human) IQ tests. This work presents a new approach to machine IQ tests solving formulated as Raven’s Progressive Matrices (RPMs), called Duel-IQ. The proposed solution incorporates the concept of a tournament in which the best answer is chosen based on a set of duels between candidate RPM answers. The three relevant aspects are: (1) low computational and design complexity, (2) proposition of two schemes of pairing up candidate answers for the duels and (3) evaluation of the system on a dataset of shapes other than those used for training. Depending on a particular variant, the system reaches up to $82.8%$ accuracy on average in RPM tasks with 5 candidate answers and is on par with human performance and superior to other literature approaches of comparable complexity when training and test sets are from the same distribution. } }
Endnote
%0 Conference Paper %T Duel-based Deep Learning system for solving IQ tests %A Paulina Tomaszewska %A Adam Żychowski %A Jacek Mańdziuk %B Proceedings of The 25th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2022 %E Gustau Camps-Valls %E Francisco J. R. Ruiz %E Isabel Valera %F pmlr-v151-tomaszewska22a %I PMLR %P 10483--10492 %U https://proceedings.mlr.press/v151/tomaszewska22a.html %V 151 %X One of the relevant aspects of Artificial General Intelligence is the ability of machines to demonstrate abstract reasoning skills, for instance, through solving (human) IQ tests. This work presents a new approach to machine IQ tests solving formulated as Raven’s Progressive Matrices (RPMs), called Duel-IQ. The proposed solution incorporates the concept of a tournament in which the best answer is chosen based on a set of duels between candidate RPM answers. The three relevant aspects are: (1) low computational and design complexity, (2) proposition of two schemes of pairing up candidate answers for the duels and (3) evaluation of the system on a dataset of shapes other than those used for training. Depending on a particular variant, the system reaches up to $82.8%$ accuracy on average in RPM tasks with 5 candidate answers and is on par with human performance and superior to other literature approaches of comparable complexity when training and test sets are from the same distribution.
APA
Tomaszewska, P., Żychowski, A. & Mańdziuk, J.. (2022). Duel-based Deep Learning system for solving IQ tests . Proceedings of The 25th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 151:10483-10492 Available from https://proceedings.mlr.press/v151/tomaszewska22a.html.

Related Material