SFFAI論壇已開放注冊

詳情點(diǎn)擊查看：https://bbs.sffai.com/d/312

深度學(xué)習(xí)和符號(hào)推理是智能系統(tǒng)中的互補(bǔ)技術(shù)，本期論壇講者黃佳妮提出了Scallop系統(tǒng)，在概率演繹數(shù)據(jù)庫上結(jié)合了兩種技術(shù)，在需要多跳推理的視覺問答任務(wù)展現(xiàn)出了獨(dú)特優(yōu)勢。

講者介紹

黃佳妮，賓夕法尼亞大學(xué)博四學(xué)生，導(dǎo)師是Mayur Naik。主要研究方向是機(jī)器學(xué)習(xí)和編程語言的交叉領(lǐng)域：運(yùn)用PL的方法，以及神經(jīng)符號(hào)方法，使學(xué)習(xí)的過程更加強(qiáng)健，數(shù)據(jù)的使用更加高效。目前在NeurIPS和ICML會(huì)議上發(fā)表論文2篇。

分享題目

Scallop: From Probabilistic Deductive Databases to Scalable Differentiable Reasoning

分享摘要

Deep learning and symbolic reasoning are complementary techniques for an intelligent system. However, principled combinations of these techniques are typically limited in scalability, rendering them ill-suited for real-world applications. We propose Scallop, a system that builds upon probabilistic deductive databases, to bridge this gap. On synthetic tasks involving mathematical and logical reasoning, Scallop scales significantly better without sacrificing accuracy compared to DeepProbLog, a principled neural logic programming approach. Scallop also scales to a real-world Visual Question Answering (VQA) benchmark that requires multi-hop reasoning, achieving 84.22% accuracy and outperforming two VQA-tailored models based on Neural Module Networks and transformers by 12.42% and 21.66% respectively

分享亮點(diǎn)

1.?We introduce the notion of top-k proofs which generalizes exact probabilistic reasoning, asymptotically reduces computational cost, and provides relative accuracy guarantees.

2.?We design and implement a framework, Scallop, which introduces a tunable parameter k and efficiently implements the computation of top-k proofs using provenance in Datalog, while retaining the benefits of neural and symbolic approaches.

3.?We empirically evaluate Scallop on synthetic tasks as well as a real-world task, VQA with multi-hop reasoning, and demonstrate that it significantly outperforms baselines.

直播時(shí)間

2022年2月20日（周日）20:00—21:00 線上直播

關(guān)注微信公眾號(hào)：人工智能前沿講習(xí)，對話框回復(fù)“SFFAI137”，獲取入群二維碼