Leveraging Language Models to Efficiently Learn Symbolic Optimization Solutions

Symbolic Optimization has been used to solve varied challenging and relevant problems such as Symbolic Regression and Neural Architecture Search. However, the current state-of-the-art typically learns each problem from scratch , and is unable to leverage pre-existing knowledge and datasets that are available for many applications. Inspired by the similarity between sequence representations learned in Natural Language Processing and the formulation of symbolic optimization as a discrete sequence optimization problem, we propose Language Model-Accelerated Deep Symbolic Optimization (LA-DSO), a method that leverages language models to learn symbolic optimization solutions more efficiently. We demonstrate LA-DSO in two tasks: symbolic regression, which allows us to com-pare against the state-of-the-art approaches, and computational antibody optimization, which shows that our proposal accelerates learning for challenging real-world problems.

[1]  S. Gu,et al.  Can Wikipedia Help Offline Reinforcement Learning? , 2022, ArXiv.

[2]  Mohammad Taher Pilehvar,et al.  Embeddings in Natural Language Processing: Theory and Advances in Vector Representations of Meaning , 2020, Embeddings in Natural Language Processing.

[3]  B. Rost,et al.  ProtTrans: Towards Cracking the Language of Life’s Code Through Self-Supervised Deep Learning and High Performance Computing , 2020, bioRxiv.

[4]  Mark Chen,et al.  Language Models are Few-Shot Learners , 2020, NeurIPS.

[5]  A. Zemla,et al.  Rapid in silico design of antibodies targeting SARS-CoV-2 using machine learning and supercomputing , 2020, bioRxiv.

[6]  Jiwei Li,et al.  Description Based Text Classification with Reinforcement Learning , 2020, ICML.

[7]  Brenden K. Petersen,et al.  Deep symbolic regression: Recovering mathematical expressions from data via risk-seeking policy gradients , 2019, ICLR.

[8]  Colin Raffel,et al.  Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer , 2019, J. Mach. Learn. Res..

[9]  Sebastian Kelm,et al.  Computational approaches to therapeutic antibody design: established methods and emerging trends , 2019, Briefings Bioinform..

[10]  Shimon Whiteson,et al.  A Survey of Reinforcement Learning Informed by Natural Language , 2019, IJCAI.

[11]  Thomas Wolf,et al.  Transfer Learning in Natural Language Processing , 2019, NAACL.

[12]  Martin Jaggi,et al.  Evaluating the Search Phase of Neural Architecture Search , 2019, ICLR.

[13]  A. Walls,et al.  Unexpected Receptor Functional Mimicry Elucidates Activation of Coronavirus Fusion , 2019, Cell.

[14]  Pushmeet Kohli,et al.  Learning to Understand Goal Specifications by Modelling Reward , 2018, ICLR.

[15]  Pushmeet Kohli,et al.  Learning to Follow Language Instructions with Adversarial Reward Induction , 2018, ArXiv.

[16]  Wei Xu,et al.  Interactive Grounded Language Acquisition and Generalization in a 2D World , 2018, ICLR.

[17]  Markus Heinonen,et al.  Flex ddG: Rosetta ensemble-based estimation of changes in protein-protein binding affinity upon mutation , 2017, bioRxiv.

[18]  Regina Barzilay,et al.  Grounding Language for Transfer in Deep Reinforcement Learning , 2017, J. Artif. Intell. Res..

[19]  Demis Hassabis,et al.  Grounded Language Learning in a Simulated 3D World , 2017, ArXiv.

[20]  Andy R. Terrel,et al.  SymPy: Symbolic computing in Python , 2017, PeerJ Prepr..

[21]  Jun Ren,et al.  Using Genetic Programming with Prior Formula Knowledge to Solve Symbolic Regression Problem , 2015, Comput. Intell. Neurosci..

[22]  Jeffrey Pennington,et al.  GloVe: Global Vectors for Word Representation , 2014, EMNLP.

[23]  Jeffrey Dean,et al.  Efficient Estimation of Word Representations in Vector Space , 2013, ICLR.

[24]  Wojciech Jaskowski,et al.  Better GP benchmarks: community survey results and proposals , 2012, Genetic Programming and Evolvable Machines.

[25]  Michael O'Neill,et al.  Genetic Programming and Evolvable Machines Manuscript No. Semantically-based Crossover in Genetic Programming: Application to Real-valued Symbolic Regression , 2022 .

[26]  Qiang Yang,et al.  A Survey on Transfer Learning , 2010, IEEE Transactions on Knowledge and Data Engineering.

[27]  Emanuel Kitzelmann,et al.  Inductive Programming: A Survey of Program Synthesis Techniques , 2009, AAIP.

[28]  D. Dimitrov,et al.  Potent cross-reactive neutralization of SARS coronavirus isolates by human monoclonal antibodies , 2007, Proceedings of the National Academy of Sciences.

[29]  P. Carter.  Potent antibody therapeutics by design , 2006, Nature Reviews Immunology.

[30]  Zoubin Ghahramani,et al.  Sparse Gaussian Processes using Pseudo-inputs , 2005, NIPS.

[31]  Wenhui Li,et al.  Potent neutralization of severe acute respiratory syndrome (SARS) coronavirus by a human mAb to S1 protein that blocks receptor association. , 2004, Proceedings of the National Academy of Sciences of the United States of America.

[32]  T. T. Wu,et al.  AN ANALYSIS OF THE SEQUENCES OF THE VARIABLE REGIONS OF BENCE JONES PROTEINS AND MYELOMA LIGHT CHAINS AND THEIR IMPLICATIONS FOR ANTIBODY COMPLEMENTARITY , 1970, The Journal of experimental medicine.

[33]  Felipe Leno da Silva,et al.  Deep Symbolic Optimization for Electric Component Sizing in Fixed Topology Power Converters , 2021 .

[34]  Ruben Glatt,et al.  Discovering symbolic policies with deep reinforcement learning , 2021, ICML.

[35]  Brenden K. Petersen,et al.  Learning sparse symbolic policies for sepsis treatment , 2021 .

[36]  Brenden K. Petersen,et al.  Symbolic Regression via Deep Reinforcement Learning Enhanced Genetic Programming Seeding , 2021, NeurIPS.

[37]  Ming-Wei Chang,et al.  BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding , 2019, NAACL.

[38]  Jens Meiler,et al.  ROSETTA3: an object-oriented software suite for the simulation and design of macromolecules. , 2011, Methods in enzymology.

[39]  Lukás Burget,et al.  Recurrent neural network based language model , 2010, INTERSPEECH.