Pub Date : 2022-07-07DOI: 10.48550/arXiv.2207.03172
Gabriele Lagani, C. Gennaro, Hannes Fassold, G. Amato
Learning algorithms for Deep Neural Networks are typically based on supervised end-to-end Stochastic Gradient Descent (SGD) training with error backpropagation (backprop). Backprop algorithms require a large number of labelled training samples to achieve high performance. However, in many realistic applications, even if there is plenty of image samples, very few of them are labelled, and semi-supervised sample-efficient training strategies have to be used. Hebbian learning represents a possible approach towards sample efficient training; however, in current solutions, it does not scale well to large datasets. In this paper, we present FastHebb, an efficient and scalable solution for Hebbian learning which achieves higher efficiency by 1) merging together update computation and aggregation over a batch of inputs, and 2) leveraging efficient matrix multiplication algorithms on GPU. We validate our approach on different computer vision benchmarks, in a semi-supervised learning scenario. FastHebb outperforms previous solutions by up to 50 times in terms of training speed, and notably, for the first time, we are able to bring Hebbian algorithms to ImageNet scale.
{"title":"FastHebb: Scaling Hebbian Training of Deep Neural Networks to ImageNet Level","authors":"Gabriele Lagani, C. Gennaro, Hannes Fassold, G. Amato","doi":"10.48550/arXiv.2207.03172","DOIUrl":"https://doi.org/10.48550/arXiv.2207.03172","url":null,"abstract":"Learning algorithms for Deep Neural Networks are typically based on supervised end-to-end Stochastic Gradient Descent (SGD) training with error backpropagation (backprop). Backprop algorithms require a large number of labelled training samples to achieve high performance. However, in many realistic applications, even if there is plenty of image samples, very few of them are labelled, and semi-supervised sample-efficient training strategies have to be used. Hebbian learning represents a possible approach towards sample efficient training; however, in current solutions, it does not scale well to large datasets. In this paper, we present FastHebb, an efficient and scalable solution for Hebbian learning which achieves higher efficiency by 1) merging together update computation and aggregation over a batch of inputs, and 2) leveraging efficient matrix multiplication algorithms on GPU. We validate our approach on different computer vision benchmarks, in a semi-supervised learning scenario. FastHebb outperforms previous solutions by up to 50 times in terms of training speed, and notably, for the first time, we are able to bring Hebbian algorithms to ImageNet scale.","PeriodicalId":90051,"journal":{"name":"Similarity search and applications : proceedings of the ... International Conference on Similarity Search and Applications","volume":"294 1","pages":"251-264"},"PeriodicalIF":0.0,"publicationDate":"2022-07-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75684465","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-04-14DOI: 10.48550/arXiv.2204.07221
Paras Sheth, Ruocheng Guo, Lu Cheng, Huan Liu, K. Candan
Recommender systems aim to recommend new items to users by learning user and item representations. In practice, these representations are highly entangled as they consist of information about multiple factors, including user's interests, item attributes along with confounding factors such as user conformity, and item popularity. Considering these entangled representations for inferring user preference may lead to biased recommendations (e.g., when the recommender model recommends popular items even if they do not align with the user's interests). Recent research proposes to debias by modeling a recommender system from a causal perspective. The exposure and the ratings are analogous to the treatment and the outcome in the causal inference framework, respectively. The critical challenge in this setting is accounting for the hidden confounders. These confounders are unobserved, making it hard to measure them. On the other hand, since these confounders affect both the exposure and the ratings, it is essential to account for them in generating debiased recommendations. To better approximate hidden confounders, we propose to leverage network information (i.e., user-social and user-item networks), which are shown to influence how users discover and interact with an item. Aside from the user conformity, aspects of confounding such as item popularity present in the network information is also captured in our method with the aid of textit{causal disentanglement} which unravels the learned representations into independent factors that are responsible for (a) modeling the exposure of an item to the user, (b) predicting the ratings, and (c) controlling the hidden confounders. Experiments on real-world datasets validate the effectiveness of the proposed model for debiasing recommender systems.
{"title":"Causal Disentanglement with Network Information for Debiased Recommendations","authors":"Paras Sheth, Ruocheng Guo, Lu Cheng, Huan Liu, K. Candan","doi":"10.48550/arXiv.2204.07221","DOIUrl":"https://doi.org/10.48550/arXiv.2204.07221","url":null,"abstract":"Recommender systems aim to recommend new items to users by learning user and item representations. In practice, these representations are highly entangled as they consist of information about multiple factors, including user's interests, item attributes along with confounding factors such as user conformity, and item popularity. Considering these entangled representations for inferring user preference may lead to biased recommendations (e.g., when the recommender model recommends popular items even if they do not align with the user's interests). Recent research proposes to debias by modeling a recommender system from a causal perspective. The exposure and the ratings are analogous to the treatment and the outcome in the causal inference framework, respectively. The critical challenge in this setting is accounting for the hidden confounders. These confounders are unobserved, making it hard to measure them. On the other hand, since these confounders affect both the exposure and the ratings, it is essential to account for them in generating debiased recommendations. To better approximate hidden confounders, we propose to leverage network information (i.e., user-social and user-item networks), which are shown to influence how users discover and interact with an item. Aside from the user conformity, aspects of confounding such as item popularity present in the network information is also captured in our method with the aid of textit{causal disentanglement} which unravels the learned representations into independent factors that are responsible for (a) modeling the exposure of an item to the user, (b) predicting the ratings, and (c) controlling the hidden confounders. Experiments on real-world datasets validate the effectiveness of the proposed model for debiasing recommender systems.","PeriodicalId":90051,"journal":{"name":"Similarity search and applications : proceedings of the ... International Conference on Similarity Search and Applications","volume":"13 1","pages":"265-273"},"PeriodicalIF":0.0,"publicationDate":"2022-04-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84855699","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-01-01DOI: 10.48550/arXiv.2208.08910
Jaroslav Olha, Terézia Slanináková, Martin Gendiar, Matej Antol, Vlastislav Dohnal
. Despite the constant evolution of similarity searching research, it continues to face the same challenges stemming from the complexity of the data, such as the curse of dimensionality and computationally expensive distance functions. Various machine learning techniques have proven capable of replacing elaborate mathematical models with combinations of simple linear functions, often gaining speed and sim-plicity at the cost of formal guarantees of accuracy and correctness of querying.Theauthors explore the potential of this research trend by presenting a lightweight solution for the complex problem of 3D protein structure search. The solution consists of three steps – (i) transformation of 3D protein structural information into very compact vectors, (ii) use of a probabilistic model to group these vectors and respond to queries by returning a given number of similar objects, and (iii) a final filtering step which applies basic vector distance functions to refine the result.
{"title":"Learned Indexing in Proteins: Substituting Complex Distance Calculations with Embedding and Clustering Techniques","authors":"Jaroslav Olha, Terézia Slanináková, Martin Gendiar, Matej Antol, Vlastislav Dohnal","doi":"10.48550/arXiv.2208.08910","DOIUrl":"https://doi.org/10.48550/arXiv.2208.08910","url":null,"abstract":". Despite the constant evolution of similarity searching research, it continues to face the same challenges stemming from the complexity of the data, such as the curse of dimensionality and computationally expensive distance functions. Various machine learning techniques have proven capable of replacing elaborate mathematical models with combinations of simple linear functions, often gaining speed and sim-plicity at the cost of formal guarantees of accuracy and correctness of querying.Theauthors explore the potential of this research trend by presenting a lightweight solution for the complex problem of 3D protein structure search. The solution consists of three steps – (i) transformation of 3D protein structural information into very compact vectors, (ii) use of a probabilistic model to group these vectors and respond to queries by returning a given number of similar objects, and (iii) a final filtering step which applies basic vector distance functions to refine the result.","PeriodicalId":90051,"journal":{"name":"Similarity search and applications : proceedings of the ... International Conference on Similarity Search and Applications","volume":"19 1","pages":"274-282"},"PeriodicalIF":0.0,"publicationDate":"2022-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82534773","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-01-01DOI: 10.1007/978-3-031-17849-8_23
G. Moro, Lorenzo Valgimigli, Alex Rossi, Cristiano Casadei, Andrea Montefiori
{"title":"Self-supervised Information Retrieval Trained from Self-generated Sets of Queries and Relevant Documents","authors":"G. Moro, Lorenzo Valgimigli, Alex Rossi, Cristiano Casadei, Andrea Montefiori","doi":"10.1007/978-3-031-17849-8_23","DOIUrl":"https://doi.org/10.1007/978-3-031-17849-8_23","url":null,"abstract":"","PeriodicalId":90051,"journal":{"name":"Similarity search and applications : proceedings of the ... International Conference on Similarity Search and Applications","volume":"26 1","pages":"283-290"},"PeriodicalIF":0.0,"publicationDate":"2022-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75413503","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-01-01DOI: 10.1007/978-3-031-17849-8_25
Alessandro Abluton
{"title":"Visual Recommendation and Visual Search for Fashion E-Commerce","authors":"Alessandro Abluton","doi":"10.1007/978-3-031-17849-8_25","DOIUrl":"https://doi.org/10.1007/978-3-031-17849-8_25","url":null,"abstract":"","PeriodicalId":90051,"journal":{"name":"Similarity search and applications : proceedings of the ... International Conference on Similarity Search and Applications","volume":"97 1","pages":"299-304"},"PeriodicalIF":0.0,"publicationDate":"2022-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79989399","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-01-01DOI: 10.1007/978-3-031-17849-8_9
Lucia Vadicamo, A. Dearle, R. Connor
{"title":"On the Expected Exclusion Power of Binary Partitions for Metric Search","authors":"Lucia Vadicamo, A. Dearle, R. Connor","doi":"10.1007/978-3-031-17849-8_9","DOIUrl":"https://doi.org/10.1007/978-3-031-17849-8_9","url":null,"abstract":"","PeriodicalId":90051,"journal":{"name":"Similarity search and applications : proceedings of the ... International Conference on Similarity Search and Applications","volume":"13 1","pages":"104-117"},"PeriodicalIF":0.0,"publicationDate":"2022-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73954380","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-01-01DOI: 10.1007/978-3-031-17849-8_8
Vladimir Mic, P. Zezula
{"title":"Concept of Relational Similarity Search","authors":"Vladimir Mic, P. Zezula","doi":"10.1007/978-3-031-17849-8_8","DOIUrl":"https://doi.org/10.1007/978-3-031-17849-8_8","url":null,"abstract":"","PeriodicalId":90051,"journal":{"name":"Similarity search and applications : proceedings of the ... International Conference on Similarity Search and Applications","volume":"1 1","pages":"89-103"},"PeriodicalIF":0.0,"publicationDate":"2022-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"72910489","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-01-01DOI: 10.1007/978-3-031-17849-8_17
F. Carrara, Lucia Vadicamo, C. Gennaro, G. Amato
{"title":"Approximate Nearest Neighbor Search on Standard Search Engines","authors":"F. Carrara, Lucia Vadicamo, C. Gennaro, G. Amato","doi":"10.1007/978-3-031-17849-8_17","DOIUrl":"https://doi.org/10.1007/978-3-031-17849-8_17","url":null,"abstract":"","PeriodicalId":90051,"journal":{"name":"Similarity search and applications : proceedings of the ... International Conference on Similarity Search and Applications","volume":"5 1","pages":"214-221"},"PeriodicalIF":0.0,"publicationDate":"2022-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"81745795","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-01-01DOI: 10.1007/978-3-031-17849-8_14
Ibrahim Chegrane, Imane Hocine, Saïd Yahiaoui, A. Bendjoudi, Nadia Nouali-Taboudjemat
{"title":"Graph Edit Distance Compacted Search Tree","authors":"Ibrahim Chegrane, Imane Hocine, Saïd Yahiaoui, A. Bendjoudi, Nadia Nouali-Taboudjemat","doi":"10.1007/978-3-031-17849-8_14","DOIUrl":"https://doi.org/10.1007/978-3-031-17849-8_14","url":null,"abstract":"","PeriodicalId":90051,"journal":{"name":"Similarity search and applications : proceedings of the ... International Conference on Similarity Search and Applications","volume":"23 1","pages":"181-189"},"PeriodicalIF":0.0,"publicationDate":"2022-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86878471","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-01-01DOI: 10.1007/978-3-031-17849-8_4
G. Moro, S. Salvatori
{"title":"Deep Vision-Language Model for Efficient Multi-modal Similarity Search in Fashion Retrieval","authors":"G. Moro, S. Salvatori","doi":"10.1007/978-3-031-17849-8_4","DOIUrl":"https://doi.org/10.1007/978-3-031-17849-8_4","url":null,"abstract":"","PeriodicalId":90051,"journal":{"name":"Similarity search and applications : proceedings of the ... International Conference on Similarity Search and Applications","volume":"29 1","pages":"40-53"},"PeriodicalIF":0.0,"publicationDate":"2022-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"81669444","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：