{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2022,11,2]],"date-time":"2022-11-02T05:07:07Z","timestamp":1667365627097},"reference-count":27,"publisher":"MIT Press","issue":"4","content-domain":{"domain":["direct.mit.edu"],"crossmark-restriction":true},"short-container-title":[],"published-print":{"date-parts":[[2022,1,1]]},"abstract":"<jats:title>Abstract<\/jats:title>\n               <jats:p>Understanding the structure and evolution of natural cognition is a topic of broad scientific interest, as is the development of an engineering toolkit to construct artificial cognitive systems. One open question is determining which components and techniques to use in such a toolkit. To investigate this question, we employ agent-based AI, using simple computational substrates (i.e., digital brains) undergoing rapid evolution. Such systems are an ideal choice as they are fast to process, easy to manipulate, and transparent for analysis. Even in this limited domain, however, hundreds of different computational substrates are used. While benchmarks exist to compare the quality of different substrates, little work has been done to build broader theory on how substrate features interact. We propose a technique called the Comparative Hybrid Approach and develop a proof-of-concept by systematically analyzing components from three evolvable substrates: recurrent artificial neural networks, Markov brains, and Cartesian genetic programming. We study the role and interaction of individual elements of these substrates by recombining them in a piecewise manner to form new hybrid substrates that can be empirically tested. Here, we focus on network sparsity, memory discretization, and logic operators of each substrate. We test the original substrates and the hybrids across a suite of distinct environments with different logic and memory requirements. While we observe many trends, we see that discreteness of memory and the Markov brain logic gates correlate with high performance across our test conditions. Our results demonstrate that the Comparative Hybrid Approach can identify structural subcomponents that predict task performance across multiple computational substrates.<\/jats:p>","DOI":"10.1162\/artl_a_00372","type":"journal-article","created":{"date-parts":[[2022,8,5]],"date-time":"2022-08-05T13:36:00Z","timestamp":1659706560000},"page":"423-439","update-policy":"http:\/\/dx.doi.org\/10.1162\/mitpressjournals.corrections.policy","source":"Crossref","is-referenced-by-count":1,"title":["Using the Comparative Hybrid Approach to Disentangle the Role of Substrate Choice on the Evolution of Cognition"],"prefix":"10.1162","volume":"28","author":[{"given":"Clifford","family":"Bohm","sequence":"first","affiliation":[{"name":"Michigan State University, Department of Integrative Biology, BEACON Center for the Study of Evolution in Action. cliff.bohm@gmail.com"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Sarah","family":"Albani","sequence":"additional","affiliation":[{"name":"Michigan State University, Department of Neuroscience, Lyman Briggs College BEACON Center for the Study of Evolution in Action"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Charles","family":"Ofria","sequence":"additional","affiliation":[{"name":"Michigan State University, Department of Computer Science Program in Ecology, Evolution, and Biology BEACON Center for the Study of Evolution in Action"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Acacia","family":"Ackles","sequence":"additional","affiliation":[{"name":"Michigan State University, Department of Integrative Biology Program in Ecology, Evolution, and Biology BEACON Center for the Study of Evolution in Action"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"281","published-online":{"date-parts":[[2022,11,1]]},"reference":[{"key":"2022110114032673700_bib1","doi-asserted-by":"publisher","first-page":"110","DOI":"10.1162\/isal_a_00453","article-title":"The comparative hybrid approach to investigate cognition across substrates","volume-title":"ALIFE 2021: The 2021 conference on artificial life","author":"Albani","year":"2021"},{"key":"2022110114032673700_bib2","doi-asserted-by":"publisher","first-page":"Article 100001","DOI":"10.1016\/j.eswax.2019.100001","article-title":"A review of machine learning algorithms for identification and classification of non-functional requirements","volume":"1","author":"Binkhonain","year":"2019","journal-title":"Expert Systems with Applications: X"},{"key":"2022110114032673700_bib3","doi-asserted-by":"publisher","first-page":"76","DOI":"10.1162\/isal_a_016","volume-title":"MABE (modular agent based evolver): A framework for digital evolution research","author":"Bohm","year":"2017"},{"key":"2022110114032673700_bib4","doi-asserted-by":"publisher","first-page":"113","DOI":"10.1016\/j.ecolmodel.2012.03.001","article-title":"A review of supervised machine learning algorithms and their applications to ecological data","volume":"240","author":"Crisci","year":"2012","journal-title":"Ecological Modelling"},{"issue":"10","key":"2022110114032673700_bib5","doi-asserted-by":"publisher","first-page":"Article e1002236","DOI":"10.1371\/journal.pcbi.1002236","article-title":"Integrated information increases with fitness in the evolution of animats","volume":"7","author":"Edlund","year":"2011","journal-title":"PLoS Computational Biology"},{"issue":"4","key":"2022110114032673700_bib6","doi-asserted-by":"publisher","first-page":"295","DOI":"10.1142\/S0129065709002002","article-title":"Spiking neural networks","volume":"19","author":"Ghosh-Dastidar","year":"2009","journal-title":"International Journal of Neural Systems"},{"key":"2022110114032673700_bib7","doi-asserted-by":"publisher","first-page":"115","DOI":"10.1109\/ISCMI51676.2020.9311592","article-title":"Cryptic information transfer in differently-trained recurrent neural networks","volume-title":"ISMCI 2020: 7th international conference on soft computing & machine intelligence","author":"Hintze","year":"2020"},{"key":"2022110114032673700_bib8","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1709.05601","volume-title":"Markov brains: A technical introduction","author":"Hintze","year":"2017"},{"key":"2022110114032673700_bib9","doi-asserted-by":"publisher","first-page":"17","DOI":"10.1007\/978-3-030-04735-1_2","article-title":"The evolutionary buffet method","volume-title":"Genetic programming theory and practice XVI","author":"Hintze","year":"2019"},{"issue":"51","key":"2022110114032673700_bib10","doi-asserted-by":"publisher","first-page":"Article 2290","DOI":"10.21105\/joss.02290","article-title":"dcgp: Differentiable Cartesian genetic programming made easy","volume":"5","author":"Izzo","year":"2020","journal-title":"Journal of Open Source Software"},{"issue":"4","key":"2022110114032673700_bib11","doi-asserted-by":"publisher","first-page":"394","DOI":"10.1080\/09658211003702171","article-title":"The concurrent validity of the N-back task as a working memory measure","volume":"18","author":"Jaeggi","year":"2010","journal-title":"Memory"},{"key":"2022110114032673700_bib12","doi-asserted-by":"publisher","first-page":"195","DOI":"10.1145\/3071178.3071303","article-title":"Multi-task learning in Atari video games with emergent tangled program graphs","volume-title":"GECCO \u201917: Proceedings of the genetic and evolutionary computation conference","author":"Kelly","year":"2017"},{"issue":"1","key":"2022110114032673700_bib13","doi-asserted-by":"publisher","first-page":"4","DOI":"10.4304\/jait.1.1.4-20","article-title":"A review of machine learning algorithms for text-documents classification","volume":"1","author":"Khan","year":"2010","journal-title":"Journal of Advances in Information Technology"},{"key":"2022110114032673700_bib14","doi-asserted-by":"publisher","first-page":"1135","DOI":"10.1145\/3205455.3205523","article-title":"Evolving event-driven programs with SignalGP","volume-title":"GECCO \u201918: Proceedings of the genetic and evolutionary computation conference","author":"Lalejini","year":"2018"},{"issue":"8","key":"2022110114032673700_bib15","doi-asserted-by":"publisher","first-page":"2079","DOI":"10.1162\/NECO_a_00475","article-title":"The evolution of representation in simple cognitive networks","volume":"25","author":"Marstaller","year":"2013","journal-title":"Neural Computation"},{"key":"2022110114032673700_bib16","first-page":"1135","article-title":"An empirical study of the efficiency of learning Boolean functions using a Cartesian genetic programming approach","volume-title":"GECCO\u201999: Proceedings of the genetic and evolutionary computation conference","author":"Miller","year":"1999"},{"issue":"7","key":"2022110114032673700_bib17","doi-asserted-by":"publisher","first-page":"711","DOI":"10.1093\/arclin\/acp063","article-title":"Is the N-Back task a valid neuropsychological measure for assessing working memory?","volume":"24","author":"Miller","year":"2009","journal-title":"Archives of Clinical Neuropsychology"},{"issue":"2","key":"2022110114032673700_bib18","doi-asserted-by":"publisher","first-page":"191","DOI":"10.1162\/106454604773563612","article-title":"Avida: A software platform for research in computational evolutionary biology","volume":"10","author":"Ofria","year":"2004","journal-title":"Artificial Life"},{"key":"2022110114032673700_bib19","doi-asserted-by":"publisher","first-page":"46","DOI":"10.1002\/hbm.20131","article-title":"N-Back working memory paradigm: A meta-analysis of normative functional neuroimaging studies","volume":"25","author":"Owen","year":"2005","journal-title":"Human Brain Mapping"},{"issue":"1","key":"2022110114032673700_bib20","doi-asserted-by":"publisher","first-page":"E1","DOI":"10.1086\/706252","article-title":"The evolutionary origin of associative learning","volume":"195","author":"Pontes","year":"2020","journal-title":"The American Naturalist"},{"key":"2022110114032673700_bib21","first-page":"8007","article-title":"AutoML-zero: Evolving machine learning algorithms from scratch","volume-title":"PMLR 119: Proceedings of the 37th international conference on machine learning","author":"Real","year":"2020"},{"issue":"6","key":"2022110114032673700_bib22","doi-asserted-by":"publisher","first-page":"386","DOI":"10.1037\/h0042519","article-title":"The perceptron: A probabilistic model for information storage and organization in the brain","volume":"65","author":"Rosenblatt","year":"1958","journal-title":"Psychological Review"},{"key":"2022110114032673700_bib23","doi-asserted-by":"publisher","first-page":"Article 16712","DOI":"10.1038\/s41598-017-16548-2","article-title":"Evolving autonomous learning in cognitive networks","volume":"7","author":"Sheneman","year":"2017","journal-title":"Scientific Reports"},{"key":"2022110114032673700_bib24","first-page":"1310","article-title":"A review of supervised machine learning algorithms","volume-title":"Proceedings of the 10th INDIACom: 2016 3rd international conference on computing for sustainable global development","author":"Singh","year":"2016"},{"issue":"1","key":"2022110114032673700_bib25","doi-asserted-by":"publisher","first-page":"Article 281","DOI":"10.1186\/s12911-019-1004-8","article-title":"Comparing different supervised machine learning algorithms for disease prediction","volume":"19","author":"Uddin","year":"2019","journal-title":"BMC Medical Informatics and Decision Making"},{"issue":"5","key":"2022110114032673700_bib26","doi-asserted-by":"publisher","first-page":"5","DOI":"10.1145\/1163593.1163596","article-title":"A preliminary performance comparison of five machine learning algorithms for practical IP traffic flow classification","volume":"36","author":"Williams","year":"2006","journal-title":"ACM SIGCOMM Computer Communication Review"},{"issue":"7","key":"2022110114032673700_bib27","doi-asserted-by":"publisher","first-page":"1235","DOI":"10.1162\/neco_a_01199","article-title":"A review of recurrent neural networks: LSTM cells and network architectures","volume":"31","author":"Yu","year":"2019","journal-title":"Neural Computation"}],"container-title":["Artificial Life"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/direct.mit.edu\/artl\/article-pdf\/28\/4\/423\/2043286\/artl_a_00372.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/direct.mit.edu\/artl\/article-pdf\/28\/4\/423\/2043286\/artl_a_00372.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2022,11,1]],"date-time":"2022-11-01T14:03:43Z","timestamp":1667311423000},"score":1,"resource":{"primary":{"URL":"https:\/\/direct.mit.edu\/artl\/article\/28\/4\/423\/112512\/Using-the-Comparative-Hybrid-Approach-to"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022,1,1]]},"references-count":27,"journal-issue":{"issue":"4","published-online":{"date-parts":[[2022,11,1]]},"published-print":{"date-parts":[[2022,1,1]]}},"URL":"https:\/\/doi.org\/10.1162\/artl_a_00372","relation":{},"ISSN":["1064-5462","1530-9185"],"issn-type":[{"value":"1064-5462","type":"print"},{"value":"1530-9185","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2022]]},"published":{"date-parts":[[2022,1,1]]}}}