{"paper":{"title":"Winning Lottery Tickets in Neural Networks via a Quantum-Inspired Classical Algorithm","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"A classical algorithm samples from a ridgelet-defined distribution to select sparse neural subnetworks in polynomial time in data dimension D.","cross_cats":["cs.LG","stat.ML"],"primary_cat":"quant-ph","authors_text":"Hayata Yamasaki, Mio Murao, Natsuto Isogai, Sho Sonoda","submitted_at":"2026-05-13T18:00:56Z","abstract_excerpt":"Quantum machine learning (QML) aims to accelerate machine learning tasks by exploiting quantum computation. Previous work studied a QML algorithm for selecting sparse subnetworks from large shallow neural networks. Instead of directly solving an optimization problem over a large-scale network, this algorithm constructs a sparse subnetwork by sampling hidden nodes from an optimized probability distribution defined using the ridgelet transform. The quantum algorithm performs this sampling in time $O(D)$ in the data dimension $D$, whereas a naive classical implementation relies on handling expone"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"We show that our algorithm runs in time O(poly(D)), thereby removing the exponential dependence on D from the previous classical approach.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The ridgelet transform defines an optimized probability distribution that admits an efficient classical sampling procedure with only polynomial dependence on the data dimension D.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"A classical polynomial-time algorithm for optimized sampling of lottery tickets in neural networks removes the exponential dependence on data dimension from prior classical approaches.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A classical algorithm samples from a ridgelet-defined distribution to select sparse neural subnetworks in polynomial time in data dimension D.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"88c6b0ff083a62cad5f6b0524b093abe171912ae2abbad588e539b882510bb62"},"source":{"id":"2605.13979","kind":"arxiv","version":1},"verdict":{"id":"06cfe008-dc72-4ad9-a09b-c29825a29f5d","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T05:50:22.825489Z","strongest_claim":"We show that our algorithm runs in time O(poly(D)), thereby removing the exponential dependence on D from the previous classical approach.","one_line_summary":"A classical polynomial-time algorithm for optimized sampling of lottery tickets in neural networks removes the exponential dependence on data dimension from prior classical approaches.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The ridgelet transform defines an optimized probability distribution that admits an efficient classical sampling procedure with only polynomial dependence on the data dimension D.","pith_extraction_headline":"A classical algorithm samples from a ridgelet-defined distribution to select sparse neural subnetworks in polynomial time in data dimension D."},"references":{"count":33,"sample":[{"doi":"","year":2016,"title":"Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding","work_id":"d79f6a26-2d92-4f7c-aea0-8aabf3b668c8","ref_index":1,"cited_arxiv_id":"1510.00149","is_internal_anchor":true},{"doi":"","year":2019,"title":"SNIP: SINGLE-SHOT NETWORK PRUNING BASED ON CONNECTION SENSITIVITY","work_id":"3560a2b3-dfa6-4c6f-b2aa-48b7fa8073e9","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2020,"title":"Hidenori Tanaka, Daniel Kunin, Daniel L. K. Yamins, and Surya Ganguli. Pruning neural networks without any data by iteratively conserving synaptic flow. InProceedings of the 34th International Confere","work_id":"7e8f9941-2c8e-41f4-b814-0f899becce46","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2020,"title":"Rigging the lottery: making all tickets winners","work_id":"fdc66ad3-07d5-4549-8aa8-a4d36d1bfc06","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2019,"title":"The lottery ticket hypothesis: Finding sparse, trainable neural networks","work_id":"fd1ceaf9-3775-4154-bf6f-c2c75641e98f","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":33,"snapshot_sha256":"19c4558597966f53219c4b6e20389055eb68d3774912bd268616d24b5094b0d5","internal_anchors":3},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}