人工智慧 | ShowMeAI資訊日報 #2022.06.25

語言: CN / TW / HK

持續創作,加速成長!這是我參與「掘金日新計劃 · 6 月更文挑戰」的第27天,點選檢視活動詳情

ShowMeAI日報系列全新升級!覆蓋AI人工智慧 工具&框架 | 專案&程式碼 | 博文&分享 | 資料&資源 | 研究&論文 等方向。點選檢視 歷史文章列表,在公眾號內訂閱話題 #ShowMeAI資訊日報,可接收每日最新推送。

1.工具&框架

工具:WarcDB - 爬蟲用SQLite資料庫格式,方便分享與查詢

tags: [資料庫,爬蟲]

'WarcDB: Web crawl data as SQLite databases. - WarcDB: Web crawl data as SQLite databases.' by Florents Tselai

GitHub: https://github.com/Florents-Tselai/WarcDB

工具:rspleeter - Rust寫的人聲與音樂分離工具

tags: [人聲分離,音樂分離,聲音分離]

'rspleeter - Rust implementation of Spleeter' by Donough Liu

GitHub: https://github.com/ldm0/rspleeter

工具:Dooit - 字元介面的待辦事項(TODO)管理工具

tags: [ToDo,待辦,工具]

'Dooit - A TUI todo manager' by Murli Tawari

GitHub: https://github.com/kraanzu/dooit

工具:mlsync - 機器學習資料同步工具

tags: [機器學習,資料同步]

'mlsync - Sync your ML data with your favorite productivity tools!' by PaletteML

GitHub: https://github.com/paletteml/mlsync

工具:Copy Translator - Rust寫的簡單、輕量、好用的劃詞翻譯軟體

tags: [DeepL,機器翻譯,rust]

Copy Translator利用DeepL翻譯,無需註冊

GitHub: https://github.com/zu1k/copy-translator

2.博文&分享

博文:機器學習博士“早知道就好了”的九種好工具

tags: [科研,機器學習,工具]

Docker、Conda、Weights and biases、MLflow、Screen、GitHub、Lucidchart、Inkscape、Streamlit

《Nine Tools I Wish I Mastered before My PhD in Machine Learning》by Aliaksei Mikhailiuk

Link: https://towardsdatascience.com/nine-tools-i-wish-i-mastered-before-my-phd-in-machine-learning-708c6dcb2fb0

3.資料&資源

資源列表:EyeGazeSurvey - 深度學習注視分析相關文獻列表

tags: [深度學習,注視分析,資源列表]

‘EyeGazeSurvey - Automatic Gaze Analysis ‘in-the-wild’: A Survey' by Shreya Ghosh

GitHub: https://github.com/i-am-shreya/Eye-Gaze-Survey

資源列表:Fake News Detection - 虛假新聞檢測

tags: [虛假新聞,檢測,資源列表]

This repo is a collection of AWESOME things about fake news detection, including papers, code, etc.' by ICTMCG

GitHub: https://github.com/ICTMCG/fake-news-detection

4.研究&論文

公眾號回覆關鍵字 日報,免費獲取整理好的6月論文合輯。

論文:MineDojo: Building Open-Ended Embodied Agents with Internet-Scale Knowledge

論文標題:MineDojo: Building Open-Ended Embodied Agents with Internet-Scale Knowledge

論文時間:17 Jun 2022

所屬領域:遊戲

對應任務:強化學習遊戲

論文地址https://arxiv.org/abs/2206.08853

程式碼實現https://github.com/MineDojo/MineDojo

論文作者:Linxi Fan, Guanzhi Wang, Yunfan Jiang, Ajay Mandlekar, Yuncong Yang, Haoyi Zhu, Andrew Tang, De-An Huang, Yuke Zhu, Anima Anandkumar

論文簡介:Autonomous agents have made great strides in specialist domains like Atari games and Go./自主代理在Atari遊戲和圍棋等專業領域取得了長足的進步。

論文摘要:Autonomous agents have made great strides in specialist domains like Atari games and Go. However, they typically learn tabula rasa in isolated environments with limited and manually conceived objectives, thus failing to generalize across a wide spectrum of tasks and capabilities. Inspired by how humans continually learn and adapt in the open world, we advocate a trinity of ingredients for building generalist agents: 1) an environment that supports a multitude of tasks and goals, 2) a large-scale database of multimodal knowledge, and 3) a flexible and scalable agent architecture. We introduce MineDojo, a new framework built on the popular Minecraft game that features a simulation suite with thousands of diverse open-ended tasks and an internet-scale knowledge base with Minecraft videos, tutorials, wiki pages, and forum discussions. Using MineDojo's data, we propose a novel agent learning algorithm that leverages large pre-trained video-language models as a learned reward function. Our agent is able to solve a variety of open-ended tasks specified in free-form language without any manually designed dense shaping reward. We open-source the simulation suite and knowledge bases (https://minedojo.org) to promote research towards the goal of generally capable embodied agents.

自主代理在像雅達利遊戲和圍棋這樣的專業領域中取得了巨大的進步。然而,它們通常是在孤立的環境中以有限的、人工設想的目標進行學習,因此無法在廣泛的任務和能力範圍內進行推廣。受到人類在開放世界中不斷學習和適應的啟發,我們主張建立三位一體的通用代理。1)一個支援多種任務和目標的環境,2)一個大規模的多模態知識資料庫,以及3)一個靈活和可擴充套件的代理架構。我們推出MineDojo,這是一個建立在流行的Minecraft遊戲基礎上的新框架,它的特點是具有數千個不同的開放式任務的模擬套件和一個具有Minecraft影片、教程、維基頁面和論壇討論的網際網路規模的知識庫。利用MineDojo的資料,我們提出了一種新穎的代理學習演算法,利用大型預訓練的影片語言模型作為學習獎勵函式。我們的代理能夠解決各種用自由形式的語言指定的開放式任務,而不需要任何人工設計的密集型獎勵。我們開源了模擬套件和知識庫(https://minedojo.org),以促進對普遍具有能力的具身代理目標的研究。

論文:Multiplying Matrices Without Multiplying

論文標題:Multiplying Matrices Without Multiplying

論文時間:21 Jun 2021

所屬領域:科學計算

對應任務:矩陣計算

論文地址https://arxiv.org/abs/2106.10860

程式碼實現https://github.com/dblalock/bolt , https://github.com/joennlae/halutmatmul , https://github.com/nlpodyssey/gomaddness

論文作者:Davis Blalock, John Guttag

論文簡介:Multiplying matrices is among the most fundamental and compute-intensive operations in machine learning./矩陣相乘是機器學習中最基本和最密集的操作之一。

論文摘要:Multiplying matrices is among the most fundamental and compute-intensive operations in machine learning. Consequently, there has been significant work on efficiently approximating matrix multiplies. We introduce a learning-based algorithm for this task that greatly outperforms existing methods. Experiments using hundreds of matrices from diverse domains show that it often runs 100× faster than exact matrix products and 10× faster than current approximate methods. In the common case that one matrix is known ahead of time, our method also has the interesting property that it requires zero multiply-adds. These results suggest that a mixture of hashing, averaging, and byte shuffling−the core operations of our method−could be a more promising building block for machine learning than the sparsified, factorized, and/or scalar quantized matrix products that have recently been the focus of substantial research and hardware investment.

矩陣相乘是機器學習中最基本的計算密集型操作之一。因此,在有效逼近矩陣乘法方面已經有了大量的工作。我們為這一任務引入了一種基於學習的演算法,其效能大大超過了現有的方法。使用來自不同領域的數百個矩陣進行的實驗表明,它的執行速度往往比精確的矩陣乘法快100倍,比目前的近似方法快10倍。在一個矩陣提前知道的常見情況下,我們的方法也有一個有趣的特性,即它無需乘法加法。這些結果表明,與最近成為大量研究和硬體投資焦點的稀疏化、因子化和/或標量化矩陣產品相比,混合雜湊、平均化和位元組混序--我們方法的核心操作--可能是一個更有前途的機器學習構建塊。

論文:Bridge-Tower: Building Bridges Between Encoders in Vision-Language Representation Learning

論文標題:Bridge-Tower: Building Bridges Between Encoders in Vision-Language Representation Learning

論文時間:17 Jun 2022

所屬領域:計算機視覺,自然語言處理

對應任務:表徵學習

論文地址https://arxiv.org/abs/2206.08657

程式碼實現https://github.com/microsoft/BridgeTower

論文作者:Xiao Xu, Chenfei Wu, Shachar Rosenman, Vasudev Lal, Nan Duan

論文簡介:Current VL models either use lightweight uni-modal encoders and learn to extract, align and fuse both modalities simultaneously in a cross-modal encoder, or feed the last-layer uni-modal features directly into the top cross-modal encoder, ignoring the semantic information at the different levels in the deep uni-modal encoders./目前的VL模型要麼使用輕量級的單模態編碼器,並學習在跨模態編碼器中同時提取、對齊和融合兩種模態,要麼將最後一層的單模態特徵直接送入頂部的跨模態編碼器,忽略了深度單模態編碼器中不同層面的語義資訊。

論文摘要:Vision-Language (VL) models with the Two-Tower architecture have dominated visual-language representation learning in recent years. Current VL models either use lightweight uni-modal encoders and learn to extract, align and fuse both modalities simultaneously in a cross-modal encoder, or feed the last-layer uni-modal features directly into the top cross-modal encoder, ignoring the semantic information at the different levels in the deep uni-modal encoders. Both approaches possibly restrict vision-language representation learning and limit model performance. In this paper, we introduce multiple bridge layers that build a connection between the top layers of uni-modal encoders and each layer of the cross-modal encoder. This enables comprehensive bottom-up interactions between visual and textual representations at different semantic levels, resulting in more effective cross-modal alignment and fusion. Our proposed Bridge-Tower, pre-trained with only 4M images, achieves state-of-the-art performance on various downstream vision-language tasks. On the VQAv2 test-std set, Bridge-Tower achieves an accuracy of 78.73%, outperforming the previous state-of-the-art METER model by 1.09% with the same pre-training data and almost no additional parameters and computational cost. Notably, when further scaling the model, Bridge-Tower achieves an accuracy of 81.15%, surpassing models that are pre-trained on orders-of-magnitude larger datasets. Code is available at https://github.com/microsoft/BridgeTower

近年來,具有雙塔結構的視覺語言(VL)模型在視覺語言表徵學習中占主導地位。目前的視覺語言模型要麼使用輕量級的單模態編碼器,並學習在跨模態編碼器中同時提取、調整和融合兩種模態,要麼將最後一層的單模態特徵直接送入頂部的跨模態編碼器中,忽略了深度單模態編碼器中不同層次的語義資訊。這兩種方法都可能限制了視覺-語言表徵的學習,並限制了模型的效能。在本文中,我們引入了多個橋樑層,在單模態編碼器的頂層和跨模態編碼器的每一層之間建立聯絡。這使得不同語義層面的視覺和文字表徵之間能夠進行全面的自下而上的互動,從而實現更有效的跨模態對齊和融合。我們提出的Bridge-Tower僅用400萬張影象進行了預訓練,在各種下游的視覺-語言任務中取得了最先進的效能。在VQAv2測試資料集上,Bridge-Tower達到了78.73%的準確率,在相同的預訓練資料和幾乎沒有額外引數和計算成本的情況下,比之前最先進的METER模型高出1.09%。值得注意的是,當進一步擴充套件該模型時,Bridge-Tower達到了81.15%的準確率,超過了在更大數量級的資料集上預訓練的模型。程式碼可在 https://github.com/microsoft/BridgeTower 獲取。

論文:Improving GAN Equilibrium by Raising Spatial Awareness

論文標題:Improving GAN Equilibrium by Raising Spatial Awareness

論文時間:CVPR 2022

所屬領域:計算機視覺

對應任務:生成對抗網路

論文地址https://arxiv.org/abs/2112.00718

程式碼實現https://github.com/genforce/eqgan

論文作者:Jianyuan Wang, Ceyuan Yang, Yinghao Xu, Yujun Shen, Hongdong Li, Bolei Zhou

論文簡介:We further propose to align the spatial awareness of G with the attention map induced from D. Through this way we effectively lessen the information gap between D and G. Extensive results show that our method pushes the two-player game in GANs closer to the equilibrium, leading to a better synthesis performance./我們進一步建議將G的空間意識與從D誘匯出的注意圖相一致。通過這種方式,我們有效地減少了D和G之間的資訊差距。廣泛的結果表明,我們的方法使GANs中的雙人遊戲更接近平衡,從而導致更好的合成效能。

論文摘要:The success of Generative Adversarial Networks (GANs) is largely built upon the adversarial training between a generator (G) and a discriminator (D). They are expected to reach a certain equilibrium where D cannot distinguish the generated images from the real ones. However, such an equilibrium is rarely achieved in practical GAN training, instead, D almost always surpasses G. We attribute one of its sources to the information asymmetry between D and G. We observe that D learns its own visual attention when determining whether an image is real or fake, but G has no explicit clue on which regions to focus on for a particular synthesis. To alleviate the issue of D dominating the competition in GANs, we aim to raise the spatial awareness of G. Randomly sampled multi-level heatmaps are encoded into the intermediate layers of G as an inductive bias. Thus G can purposefully improve the synthesis of certain image regions. We further propose to align the spatial awareness of G with the attention map induced from D. Through this way we effectively lessen the information gap between D and G. Extensive results show that our method pushes the two-player game in GANs closer to the equilibrium, leading to a better synthesis performance. As a byproduct, the introduced spatial awareness facilitates interactive editing over the output synthesis. Demo video and code are available at https://genforce.github.io/eqgan-sa/

生成對抗網路(GANs)的成功主要是建立在生成器(G)和判別器(D)之間的對抗性訓練上。它們被期望達到某種平衡,即D不能區分生成的影象和真實的影象。然而,這樣的平衡在實際的GAN訓練中很少實現,相反,D幾乎總是超過G。我們將其來源之一歸結為D和G之間的資訊不對稱。我們觀察到,在確定影象是真的還是假的時候,D會學習自己的視覺注意力,但G沒有明確的線索來關注特定合成的區域。為了緩解D在GANs中主導競爭的問題,我們旨在提高G的空間意識。隨機取樣的多級熱圖被編碼到G的中間層,作為一種歸納的偏向。因此,G可以有目的地改善某些影象區域的合成。我們進一步建議將G的空間意識與從D誘匯出來的注意圖相一致。通過這種方式,我們有效地減少了D和G之間的資訊差距。廣泛的結果表明,我們的方法使GANs中的雙人遊戲更接近平衡,導致更好的合成效能。作為一個副產品,引入的空間意識促進了對輸出合成的互動式編輯。演示影片和程式碼可在 https://genforce.github.io/eqgan-sa/

論文:Powershap: A Power-full Shapley Feature Selection Method

論文標題:Powershap: A Power-full Shapley Feature Selection Method

論文時間:16 Jun 2022

所屬領域:機器學習

對應任務:特徵選擇

論文地址https://arxiv.org/abs/2206.08394

程式碼實現https://github.com/predict-idlab/powershap

論文作者:Jarne Verhaeghe, Jeroen Van Der Donckt, Femke Ongenae, Sofie Van Hoecke

論文簡介:Benchmarks and simulations show that powershap outperforms other filter methods with predictive performances on par with wrapper methods while being significantly faster, often even reaching half or a third of the execution time./基準測試和模擬表明,powershap的預測效能優於其他filter方法,與包裝方法相當,同時速度明顯加快,甚至經常達到執行時間的一半或三分之一。

論文摘要:Feature selection is a crucial step in developing robust and powerful machine learning models. Feature selection techniques can be divided into two categories: filter and wrapper methods. While wrapper methods commonly result in strong predictive performances, they suffer from a large computational complexity and therefore take a significant amount of time to complete, especially when dealing with high-dimensional feature sets. Alternatively, filter methods are considerably faster, but suffer from several other disadvantages, such as (i) requiring a threshold value, (ii) not taking into account intercorrelation between features, and (iii) ignoring feature interactions with the model. To this end, we present powershap, a novel wrapper feature selection method, which leverages statistical hypothesis testing and power calculations in combination with Shapley values for quick and intuitive feature selection. Powershap is built on the core assumption that an informative feature will have a larger impact on the prediction compared to a known random feature. Benchmarks and simulations show that powershap outperforms other filter methods with predictive performances on par with wrapper methods while being significantly faster, often even reaching half or a third of the execution time. As such, powershap provides a competitive and quick algorithm that can be used by various models in different domains. Furthermore, powershap is implemented as a plug-and-play and open-source sklearn component, enabling easy integration in conventional data science pipelines. User experience is even further enhanced by also providing an automatic mode that automatically tunes the hyper-parameters of the powershap algorithm, allowing to use the algorithm without any configuration needed.

特徵選擇是開發穩健和強大的機器學習模型的一個關鍵步驟。特徵選擇技術可以分為兩類:過濾式和包裹式方法。雖然包裹式方法通常會產生強大的預測效能,但它們有很大的計算複雜性,因此需要大量的時間來完成,特別是在處理高維特徵集時。另外,過濾式方法的速度要快得多,但也有其他一些缺點,如(i)需要一個閾值,(ii)沒有考慮到特徵之間的相互關係,以及(iii)忽略了特徵與模型的相互作用。為此,我們提出了powershap,一種新穎的包裹式特徵選擇方法,它利用統計假設檢驗和功率計算與Shapley值相結合,實現快速和直觀的特徵選擇。Powershap建立在這樣一個核心假設上:與已知的隨機特徵相比,資訊量大的特徵對預測的影響更大。基準測試和模擬表明,powershap的預測效能優於其他過濾式方法,與包裹式方法相當,同時速度明顯加快,甚至經常達到執行時間的一半或三分之一。因此,powershap提供了一種有競爭力的快速演算法,可以被不同領域的各種模型所使用。此外,powershap是作為一個即插即用的開源sklearn元件實現的,可以輕鬆地整合到傳統的資料科學管道中。通過提供自動模式,自動調整powershap演算法的超引數,允許使用該演算法而不需要任何配置,使用者體驗得到進一步增強。

論文:MCUNetV2: Memory-Efficient Patch-based Inference for Tiny Deep Learning

論文標題:MCUNetV2: Memory-Efficient Patch-based Inference for Tiny Deep Learning

論文時間:28 Oct 2021

所屬領域:自然語言處理

對應任務:Image Classification,Neural Architecture Search,object-detection,Object Detection,影象分類,神經結構搜尋,物體檢測,目標檢測

論文地址https://arxiv.org/abs/2110.15352

程式碼實現https://github.com/mit-han-lab/mcunet

論文作者:Ji Lin, Wei-Ming Chen, Han Cai, Chuang Gan, Song Han

論文簡介:We further propose network redistribution to shift the receptive field and FLOPs to the later stage and reduce the computation overhead./我們進一步提出了網路重新分配,將接受場和FLOPs轉移到後期,並減少計算開銷。

論文摘要:Tiny deep learning on microcontroller units (MCUs) is challenging due to the limited memory size. We find that the memory bottleneck is due to the imbalanced memory distribution in convolutional neural network (CNN) designs: the first several blocks have an order of magnitude larger memory usage than the rest of the network. To alleviate this issue, we propose a generic patch-by-patch inference scheduling, which operates only on a small spatial region of the feature map and significantly cuts down the peak memory. However, naive implementation brings overlapping patches and computation overhead. We further propose network redistribution to shift the receptive field and FLOPs to the later stage and reduce the computation overhead. Manually redistributing the receptive field is difficult. We automate the process with neural architecture search to jointly optimize the neural architecture and inference scheduling, leading to MCUNetV2. Patch-based inference effectively reduces the peak memory usage of existing networks by 4-8x. Co-designed with neural networks, MCUNetV2 sets a record ImageNet accuracy on MCU (71.8%), and achieves >90% accuracy on the visual wake words dataset under only 32kB SRAM. MCUNetV2 also unblocks object detection on tiny devices, achieving 16.9% higher mAP on Pascal VOC compared to the state-of-the-art result. Our study largely addressed the memory bottleneck in tinyML and paved the way for various vision applications beyond image classification.

由於記憶體大小有限,在微控制器單元(MCU)上進行微小的深度學習具有挑戰性。我們發現,記憶體瓶頸是由於卷積神經網路(CNN)設計中不平衡的記憶體分佈造成的:前幾個塊的記憶體用量比網路的其他部分大一個數量級。為了緩解這個問題,我們提出了一個通用的逐塊推理排程,它只對特徵圖的一個小空間區域進行操作,並大大減少了峰值記憶體。然而,天真的實施帶來了重疊的補丁和計算的開銷。我們進一步提出了網路再分配,將接受區和FLOPs轉移到後期,減少計算開銷。手動重新分配接受區是很困難的。我們用神經結構搜尋將這一過程自動化,以共同優化神經結構和推理排程,從而形成MCUNetV2。基於補丁的推理有效地將現有網路的峰值記憶體使用率降低了4-8倍。與神經網路共同設計的MCUNetV2在MCU上創造了ImageNet準確率的記錄(71.8%),並在僅32kB的SRAM下對視覺喚醒詞資料集實現了大於90%的準確率。MCUNetV2還解除了微小裝置上的物體檢測,與最先進的結果相比,在Pascal VOC上實現了16.9%的mAP。我們的研究在很大程度上解決了tinyML的記憶體瓶頸問題,併為影象分類以外的各種視覺應用鋪平了道路。

論文:Salient Object Detection via Integrity Learning

論文標題:Salient Object Detection via Integrity Learning

論文時間:19 Jan 2021

所屬領域:計算機視覺

對應任務:object-detection,Object Detection,Salient Object Detection,物體檢測,傾斜物體檢測

論文地址https://arxiv.org/abs/2101.07663

程式碼實現https://github.com/mczhuge/ICON , https://github.com/mczhuge/Kaleido-BERT , https://github.com/mczhuge/SOCToolbox

論文作者:Mingchen Zhuge, Deng-Ping Fan, Nian Liu, Dingwen Zhang, Dong Xu, Ling Shao

論文簡介:We define the concept of integrity at both a micro and macro level./我們在微觀和巨集觀層面上定義了完整性的概念。

論文摘要:Although current salient object detection (SOD) works have achieved significant progress, they are limited when it comes to the integrity of the predicted salient regions. We define the concept of integrity at both a micro and macro level. Specifically, at the micro level, the model should highlight all parts that belong to a certain salient object. Meanwhile, at the macro level, the model needs to discover all salient objects in a given image. To facilitate integrity learning for SOD, we design a novel Integrity Cognition Network (ICON), which explores three important components for learning strong integrity features. 1) Unlike existing models, which focus more on feature discriminability, we introduce a diverse feature aggregation (DFA) component to aggregate features with various receptive fields (i.e., kernel shape and context) and increase feature diversity. Such diversity is the foundation for mining the integral salient objects. 2) Based on the DFA features, we introduce an integrity channel enhancement (ICE) component with the goal of enhancing feature channels that highlight the integral salient objects, while suppressing the other distracting ones. 3) After extracting the enhanced features, the part-whole verification (PWV) method is employed to determine whether the part and whole object features have strong agreement. Such part-whole agreements can further improve the micro-level integrity for each salient object. To demonstrate the effectiveness of our ICON, comprehensive experiments are conducted on seven challenging benchmarks. Our ICON outperforms the baseline methods in terms of a wide range of metrics. Notably, our ICON achieves about 10% relative improvement over the previous best model in terms of average false negative ratio (FNR), on six datasets. Codes and results are available at: https://github.com/mczhuge/ICON

儘管目前的突出物體檢測(SOD)工作已經取得了重大進展,但當涉及到預測突出區域的完整性時,它們是有限的。我們在微觀和巨集觀層面上都定義了完整性的概念。具體來說,在微觀層面上,模型應該突出屬於某個突出物件的所有部分。同時,在巨集觀層面上,模型需要發現給定影象中的所有突出物件。為了促進SOD的完整性學習,我們設計了一個新穎的完整性認知網路(ICON),它為學習強大的完整性特徵探索了三個重要組成部分。1)與現有的模型不同,它更注重特徵的可辨別性,我們引入了多樣化的特徵聚合(DFA)元件,以聚合具有不同感受野(即核心形狀和背景)的特徵,增加特徵的多樣性。這種多樣性是挖掘整體突出物件的基礎。2)在DFA特徵的基礎上,我們引入了完整性通道增強(ICE)元件,目的是增強突出整體突出物件的特徵通道,同時抑制其他分散注意力的特徵。3)在提取了增強的特徵後,採用部分-整體驗證(PWV)方法來確定部分和整體物件的特徵是否有強烈的一致性。這種部分-整體的一致可以進一步提高每個突出物件的微觀層面的完整性。為了證明我們的ICON的有效性,我們在七個具有挑戰性的基準上進行了綜合實驗。我們的ICON在廣泛的指標方面超過了基線方法。值得注意的是,在六個資料集上,我們的ICON在平均假陰性率(FNR)方面比以前的最佳模型取得了大約10%的相對改進。程式碼和結果見:https://github.com/mczhuge/ICON

論文:The Shapley Value in Machine Learning

論文標題:The Shapley Value in Machine Learning

論文時間:11 Feb 2022

所屬領域:機器學習

對應任務:Ensemble Pruning,feature selection,Multi-agent Reinforcement Learning,reinforcement-learning,整合剪枝,特徵選擇,多代理強化學習,強化學習

論文地址https://arxiv.org/abs/2202.05594

程式碼實現https://github.com/benedekrozemberczki/shapley , https://github.com/AstraZeneca/awesome-shapley-value

論文作者:Benedek Rozemberczki, Lauren Watson, Péter Bayer, Hao-Tsung Yang, Olivér Kiss, Sebastian Nilsson, Rik Sarkar

論文簡介:Over the last few years, the Shapley value, a solution concept from cooperative game theory, has found numerous applications in machine learning./在過去的幾年裡,Shapley值,一個來自合作博弈理論的解決概念,已經在機器學習中找到了許多應用。

論文摘要:Over the last few years, the Shapley value, a solution concept from cooperative game theory, has found numerous applications in machine learning. In this paper, we first discuss fundamental concepts of cooperative game theory and axiomatic properties of the Shapley value. Then we give an overview of the most important applications of the Shapley value in machine learning: feature selection, explainability, multi-agent reinforcement learning, ensemble pruning, and data valuation. We examine the most crucial limitations of the Shapley value and point out directions for future research.

在過去的幾年裡,Shapley值,一個來自合作博弈理論的解決方案的概念,在機器學習中發現了許多應用。在本文中,我們首先討論了合作博弈論的基本概念和Shapley值的公理特性。然後,我們概述了Shapley值在機器學習中最重要的應用:特徵選擇、可解釋性、多代理強化學習、整合剪枝和資料評估。我們研究了Shapley值的最關鍵的侷限性,並指出了未來的研究方向。

我們是 ShowMeAI,致力於傳播AI優質內容,分享行業解決方案,用知識加速每一次技術成長!點選檢視 歷史文章列表,在公眾號內訂閱話題 #ShowMeAI資訊日報,可接收每日最新推送。點選 專題合輯&電子月刊 快速瀏覽各專題全集。

「其他文章」