Filters








1,214 Hits in 3.5 sec

Using classifier cascades for scalable e-mail classification

Jay Pujara, Hal Daumé, Lise Getoor
2011 Proceedings of the 8th Annual Collaboration, Electronic messaging, Anti-Abuse and Spam Conference on - CEAS '11  
Using this method, we learn a relationship between feature costs and label hierarchies, for granular classification and cost budgets, for load-sensitive classification.  ...  E-mail classification is an area where accurate and timely results require such a trade-off. We identify two scenarios where intelligent feature acquisition can improve classifier performance.  ...  CONCLUSION The use of classifier cascades for classifying e-mail messages shows great promise.  ... 
doi:10.1145/2030376.2030383 dblp:conf/ceas/PujaraDG11 fatcat:3whipobtbzge7dcoomsbqyjr4e

Computer Vision Platform Design with MEAN Stack Basis
MEAN Stack 기반의 컴퓨터 비전 플랫폼 설계

Seonhack Hong, Kyungsoon Cho, Jinseob Yun
2015 Journal of the Korea Society of Digital Industry and Information Management  
Especially we used MongoDB for developing the performance of vision platform because the MongoDB is more akin to working with objects in a programming language than what we know of as a database.  ...  ability with Bluetooth technology for the purpose of making Android Mobile devices interface.  ...  In this paper, we used the Raspberry Pi 2 Model and PI camera module, with at least an 16 Cascade Classification In this paper, we used the Haar Logical Platform We used the MongoDB which is  ... 
doi:10.17662/ksdim.2015.11.3.001 fatcat:cm56hcaobzdede6mvhwyr5ktae

Automatically tagging email by leveraging other users' folders

Yehuda Koren, Edo Liberty, Yoelle Maarek, Roman Sandler
2011 Proceedings of the 17th ACM SIGKDD international conference on Knowledge discovery and data mining - KDD '11  
We design a novel cascade classification approach, which copes with the severe scalability and accuracy constraints we are facing.  ...  Significant efficiency gains are achieved by working within a low dimensional latent space, and by using a novel hierarchical classifier.  ...  mail engineering team for their great help in facilitating this research and pushing it into production, with special thanks to Woody Anderson, Vishwanath Ramarao, and Andrew Sloane.  ... 
doi:10.1145/2020408.2020560 dblp:conf/kdd/KorenLMS11 fatcat:ujibsh42zzgjjc3qrtlbo6gvmy

Beyond 5G: Leveraging Cell Free TDD Massive MIMO using Cascaded Deep learning [article]

Navaneet Athreya, Vishnu Raj, Sheetal Kalyani
2020 arXiv   pre-print
The proposed method is easily scalable and removes the need for relative reciprocity calibration based on the cooperation of antennas, which usually introduces dependency in Cell Free Massive MIMO systems  ...  We propose a cascade of two Deep Neural Networks (DNN) to achieve the objective.  ...  Radhakrishna Ganti, Associate Professor of Electrical Engineering, IIT Madras for useful discussion.  ... 
arXiv:1910.05705v2 fatcat:nmxhaltkpzdwdgrycy5s25ydpm

Corporate IT-support Help-Desk Process Hybrid-Automation Solution with Machine Learning Approach [article]

Kuruparan Shanmugalingam, Nisal Chandrasekara, Calvin Hindle, Gihan Fernando, Chanaka Gunawardhana
2019 arXiv   pre-print
XGBoost cascaded hierarchical models, Bi-LSTM model with word embeddings perform well showing 77.3 overall accuracy For the real world corporate email data set.  ...  Email topic modelling with various machine learning, deep-learning approaches are compared with different features for a scalable, generalised solution along with sure-shot static rules.  ...  E. Threshold Selection In order to classify only higher confident emails, the thresholds for each and every 73 categories are defined.  ... 
arXiv:1909.09018v1 fatcat:pntnsjrylrdktclpawgdvlftf4

How Many Folders Do You Really Need?

Mihajlo Grbovic, Guy Halawi, Zohar Karnin, Yoelle Maarek
2014 Proceedings of the 23rd ACM International Conference on Conference on Information and Knowledge Management - CIKM '14  
Email classification is still a mostly manual task. Consequently, most Web mail users never define a single folder.  ...  Recently however, automatic classification offering the same categories to all users has started to appear in some Web mail clients, such as AOL or Gmail.  ...  Finally this work would not have been possible without the constant support of the Mail engineering and product teams at Yahoo.  ... 
doi:10.1145/2661829.2662018 dblp:conf/cikm/GrbovicHKM14 fatcat:po36kz252jetvcjjfpjivohxuu

Deep Learning Hyper-Parameter Optimization for Video Analytics in Clouds

Muhammad Usman Yaseen, Ashiq Anjum, Omer Rana, Nikolaos Antonopoulos
2018 IEEE Transactions on Systems, Man & Cybernetics. Systems  
The system proved to be scalable, robust and customizable for a variety of different applications.  ...  Videos are fetched from cloud storage, pre-processed and a model for supporting classification is developed on these video streams using cloud-based infrastructure.  ...  Haar cascade classifier [8] is used for the detection of objects from video frames.  ... 
doi:10.1109/tsmc.2018.2840341 fatcat:y6fp7xy6ufen5pae4qjctuc6au

Heart Disease Prediction System Using Supervised Learning Classifier

Chitra R
2013 Bonfring International Journal of Software Engineering and Soft Computing  
The information in the patient record is classified using a Cascaded Neural Network (CNN) classifier.  ...  The proposed system will provide an aid for the physicians to diagnosis the disease in a more efficient way. The efficiency of the classifier is tested using the records collected from 270 patients.  ...  The SVM classifier with RBF kernel is used for classification.  ... 
doi:10.9756/bijsesc.4336 fatcat:xux47djj2jgntmiq36edjfxx7i

Efficient model sharing for scalable collaborative classification

Odysseas Papapetrou, Wolf Siberski, Stefan Siersdorfer
2014 Peer-to-Peer Networking and Applications  
To this end, we distribute locally built classification models in a network of participating users, and combine the shared classifiers into more powerful meta models.  ...  In our experiments on four large standard collections for text classification we study the resulting tradeoffs between network cost and classification accuracy.  ...  In particular, the computational cost for classifying an item e with feature vector e, using |M i | different models is O(|M i | × |e|), while the cost for evaluating it using the meta classifier, as we  ... 
doi:10.1007/s12083-014-0259-1 fatcat:5vn5emxi3bde3hc276mknrueoy

A General Active-Learning Framework for On-Road Vehicle Recognition and Tracking

Sayanan Sivaraman, Mohan Manubhai Trivedi
2010 IEEE transactions on intelligent transportation systems (Print)  
Using the query and archiving interface for active learning (QUAIL), the passively trained vehicle-recognition system is evaluated on an independent real-world data set, and informative samples are queried  ...  A passively trained recognition system is built using conventional supervised learning.  ...  E. Murphy-Chutorian for their valuable contributions to testbed design and data collection. The authors would also like to thank the associate editor and reviewers for their valuable comments.  ... 
doi:10.1109/tits.2010.2040177 fatcat:ulb2peya3zdpzlu2ztu7m34kei

Design and Implementation of a Highly Scalable, Low-Cost Distributed Traffic Violation Enforcement System in Phuket, Thailand

Somphop Limsoonthrakul, Matthew N. Dailey, Ramesh Marikhu, Vasan Timtong, Aphinya Chairat, Anant Suphavilai, Wiwat Seetamanotch, Mongkol Ekpanyapong
2021 Sustainability  
computer vision methods and another using deep learning techniques.  ...  Information technology solutions have emerged for automated traffic enforcement systems in the last decade.  ...  We adopt different detection algorithms for the two versions deployed thus far: • Haar AdaBoost cascade classifier: This version uses the classic AdaBoost cascade classifier using Haar-like features.  ... 
doi:10.3390/su13031210 fatcat:rdxumhp4sbb5nkpv2kivzgbcua

Characterizing Individual Communication Patterns [article]

R. Dean Malmgren, Jake M. Hofman, Luis A. N. Amaral, Duncan J. Watts
2009 arXiv   pre-print
We conclude that communication patterns may prove useful as an additional class of attribute data, complementing demographic and network data, for user classification and outlier detection--a point that  ...  The increasing availability of electronic communication data, such as that arising from e-mail exchange, presents social and information scientists with new possibilities for characterizing individual  ...  EMS User 1962, for example, only sent 5 e-mails while receiving 2,284 e-mails and KW User 3069337 only sent 1 e-mail while receiving 26,230 e-mails.  ... 
arXiv:0905.0106v1 fatcat:6fhs7ozf45clxez6fxhd4sesfi

Characterizing individual communication patterns

R. Dean Malmgren, Jake M. Hofman, Luis A.N. Amaral, Duncan J. Watts
2009 Proceedings of the 15th ACM SIGKDD international conference on Knowledge discovery and data mining - KDD '09  
We conclude that communication patterns may prove useful as an additional class of attribute data, complementing demographic and network data, for user classification and outlier detection-a point that  ...  The increasing availability of electronic communication data, such as that arising from e-mail exchange, presents social and information scientists with new possibilities for characterizing individual  ...  EMS User 1962, for example, only sent 5 e-mails while receiving 2,284 e-mails and KW User 3069337 only sent 1 e-mail while receiving 26,230 e-mails.  ... 
doi:10.1145/1557019.1557088 dblp:conf/kdd/MalmgrenHAW09 fatcat:fjzoyekvyjc4xmmpzyhowjbywe

A High Speed Reconfigurable Face Detection Architecture Based on AdaBoost Cascade Algorithm

Weina ZHOU, Lin DAI, Yao ZOU, Xiaoyang ZENG, Jun HAN
2012 IEICE transactions on information and systems  
This user adjustable mode makes the reconfiguration simple and efficient, and is especially suitable for portable mobile terminals whose working condition often changes frequently.  ...  . † † The author is with Shanghai Maritime University, Shanghai, 200135, China. a) E-mail: wnzhou@shmtu.edu.cn b) E-mail: junhan@fudan.edu.cn (correspondence author) DOI: 10.1587/transinf.E95.D.383 Fig  ...  In considering that, detection using classifiers fewer than 12 stages has an unacceptable FAR(False Accept Rate) for common use, they are not included in the adjustable range.  ... 
doi:10.1587/transinf.e95.d.383 fatcat:qnywjrdx3zgmfgjneygmplos4a

Asymmetric support vector machines

Shan-Hung Wu, Keng-Pei Lin, Chung-Min Chen, Ming-Syan Chen
2008 Proceeding of the 14th ACM SIGKDD international conference on Knowledge discovery and data mining - KDD 08  
Many practical applications of classification require the classifier to produce a very low false-positive rate.  ...  Such a new objective formulation allows us to raise the confidence in predicting the positives, and therefore obtain a lower chance of false-positives.  ...  For example, users are unlikely to accept a spam filter capable of identifying 1 0 0 % of spam but half of the spam predictions are actually good mails.  ... 
doi:10.1145/1401890.1401980 dblp:conf/kdd/WuLCC08 fatcat:2yqz7emqpbeonblljvuqizndp4
« Previous Showing results 1 — 15 out of 1,214 results