Backdoor Attacks against Learning-Based Algorithms

Name: Backdoor Attacks against Learning-Based Algorithms
Author: Xuemin (Sherman) Shen
ISBN: 9783031573880

Xuemin (Sherman) Shen

Language: English

Published by Springer, 2024

3031573889 / 9783031573880

Hardcover
New

See all details

Seller: AHA-BUCH GmbH, Einbeck, GermanyAHA-BUCH GmbH

5-star seller

AbeBooks seller since August 14, 2006

View this seller's items

Hardcover

Condition: New

US$ 263.90

US$ 35.03 shipping

Ships from Germany to U.S.A.

Quantity: 1 available

Add to basket

Free 30-day returns

Save for laterCreate a Want for similar items

Item description from seller

Druck auf Anfrage Neuware - Printed after ordering - This book introduces a new type of data poisoning attack, dubbed, backdoor attack. In backdoor attacks, an attacker can train the model with poisoned data to obtain a model that performs well on a normal input but behaves wrongly with crafted triggers. Backdoor attacks can occur in many scenarios where the training process is not entirely controlled, such as using third-party datasets, third-party platforms for training, or directly calling models provided by third parties. Due to the enormous threat that backdoor attacks pose to model supply chain security, they have received widespread attention from academia and industry. This book focuses on exploiting backdoor attacks in the three types of DNN applications, which are image classification, natural language processing, and federated learning.Based on the observation that DNN models are vulnerable to small perturbations, this book demonstrates that steganography and regularization can be adopted to enhance the invisibility of backdoor triggers. Based on image similarity measurement, this book presents two metrics to quantitatively measure the invisibility of backdoor triggers. The invisible trigger design scheme introduced in this book achieves a balance between the invisibility and the effectiveness of backdoor attacks. In the natural language processing domain, it is difficult to design and insert a general backdoor in a manner imperceptible to humans. Any corruption to the textual data (e.g., misspelled words or randomly inserted trigger words/sentences) must retain context-awareness and readability to human inspectors. This book introduces two novel hidden backdoor attacks, targeting three major natural language processing tasks, including toxic comment detection, neural machine translation, and question answering, depending on whether the targeted NLP platform accepts raw Unicode characters.The emerged distributed training framework, i.e., federated learning, has advantages in preserving users' privacy. It has been widely used in electronic medical applications, however, it also faced threats derived from backdoor attacks. This book presents a novel backdoor detection framework in FL-based e-Health systems. We hope this book can provide insightful lights on understanding the backdoor attacks in different types of learning-based algorithms, including computer vision, natural language processing, and federated learning. The systematic principle in this book also offers valuable guidance on the defense of backdoor attacks against future learning-based algorithms.…

Seller Inventory # 9783031573880

Contact seller about this item

Report this item

Title: Backdoor Attacks against Learning-Based Algorithms
Author: Xuemin (Sherman) Shen
Publisher: Springer
Publication year: 2024
Condition: Neu
Binding: Buch
Language: English
ISBN 10: 3031573889
ISBN 13: 9783031573880
Item weight: 424 grams
Dimensions: 241x160x15 mm

This book introduces a new type of data poisoning attack, dubbed, backdoor attack. In backdoor attacks, an attacker can train the model with poisoned data to obtain a model that performs well on a normal input but behaves wrongly with crafted triggers. Backdoor attacks can occur in many scenarios where the training process is not entirely controlled, such as using third-party datasets, third-party platforms for training, or directly calling models provided by third parties. Due to the enormous threat that backdoor attacks pose to model supply chain security, they have received widespread attention from academia and industry. This book focuses on exploiting backdoor attacks in the three types of DNN applications, which are image classification, natural language processing, and federated learning.

Based on the observation that DNN models are vulnerable to small perturbations, this book demonstrates that steganography and regularization can be adopted to enhance the invisibility of backdoor triggers. Based on image similarity measurement, this book presents two metrics to quantitatively measure the invisibility of backdoor triggers. The invisible trigger design scheme introduced in this book achieves a balance between the invisibility and the effectiveness of backdoor attacks. In the natural language processing domain, it is difficult to design and insert a general backdoor in a manner imperceptible to humans. Any corruption to the textual data (e.g., misspelled words or randomly inserted trigger words/sentences) must retain context-awareness and readability to human inspectors. This book introduces two novel hidden backdoor attacks, targeting three major natural language processing tasks, including toxic comment detection, neural machine translation, and question answering, depending on whether the targeted NLP platform accepts raw Unicode characters.

The emerged distributed training framework, i.e., federated learning, has advantages in preserving users' privacy. It has been widely used in electronic medical applications, however, it also faced threats derived from backdoor attacks. This book presents a novel backdoor detection framework in FL-based e-Health systems. We hope this book can provide insightful lights on understanding the backdoor attacks in different types of learning-based algorithms, including computer vision, natural language processing, and federated learning. The systematic principle in this book also offers valuable guidance on the defense of backdoor attacks against future learning-based algorithms.

"Synopsis" may belong to another edition of this title.

About the Author

Shaofeng Li received the B.E. degree in Software Engineering from Hunan University, China, and the M.E. degree in Computer Science from Northeastern University, China, in 2014 and 2017, respectively. He received the Ph.D. degree in Computer Science from Shanghai Jiao Tong University, Canada, in 2022. Starting from 2022, he works as a Post-doctoral fellow with the Department of Mathematics and Theory, Peng Cheng Laboratory. He focuses primarily on the areas of machine learning and security, specifically exploring the robustness of machine learning models against various adversarial attacks. His work has received the ACM CCS Best Paper Award Runner-Up.

Haojin Zhu is a Professor with Department of Computer Science and Engineering, Shanghai Jiao Tong University, China. He received his B.Sc. degree (2002) from Wuhan University (China), M.Sc.(2005) degree from Shanghai Jiao Tong University (China), both in computer science and the Ph.D. in Electrical and Computer Engineering from the University of Waterloo (Canada), in 2009. He has published in more than 60 journals, including: JSAC, TDSC, TPDS, TMC, TIFS, TWC, TVT and more than 90 international conference papers, including IEEE S&P, ACM CCS, USENIX Security, ACM MOBICOM, NDSS, ACM MOBIHOC, IEEE INFOCOM, IEEE ICDCS. He received IEEE Fellow (2023), IEEE VTS Distinguished Lecturer (2022), the IEEE ComSoc Asia-Pacific Outstanding Young Researcher Award (2014) for the contribution to wireless network security and privacy, Top 100 Most Cited Chinese Papers Published in International Journals of 2014, Supervisor of Shanghai Excellent Master Thesis, and best paper awards of IEEE ICC 2007, Chinacom 2008 and best paper award runner up for Globecom 2014, WASA 2017, and ACM CCS 2021. He is leading the Network Security and Privacy Protection (NSEC) Lab.

Wen Wu received the B.E. degree in Information Engineering from South China University of Technology, Guangzhou, China, andthe M.E. degree in Electrical Engineering from University of Science and Technology of China, Hefei, China, in 2012 and 2015, respectively. He received the Ph.D. degree in Electrical and Computer Engineering from University of Waterloo, Waterloo, ON, Canada, in 2019. Starting from 2019, he works as a Post-doctoral fellow with the Department of Electrical and Computer Engineering, University of Waterloo. Currently, he is an associate professor with the Department of Mathematics and Theory, Pengcheng Laboratory. His research interests include millimeter-wave networks and AI-empowered wireless networks.

Xuemin (Sherman) Shen received the Ph.D. degree in electrical engineering from Rutgers University, New Brunswick, NJ, USA, in 1990. He is a University Professor with the Department of Electrical and Computer Engineering, University of Waterloo, Canada. His research focuses on network resource management, wireless network security, Internet of Things, 5G andbeyond, and vehicular ad hoc and sensor networks. Dr. Shen is a registered Professional Engineer of Ontario, Canada, an Engineering Institute of Canada Fellow, a Canadian Academy of Engineering Fellow, a Royal Society of Canada Fellow, a Chinese Academy of Engineering Foreign Member, and a Distinguished Lecturer of the IEEE Vehicular Technology Society and Communications Society. Dr. Shen received the Canadian Award for Telecommunications Research from the Canadian Society of Information Theory (CSIT) in 2021, the R.A. Fessenden Award in 2019 from IEEE, Canada, Award of Merit from the Federation of Chinese Canadian Professionals (Ontario) in 2019, James Evans Avant Garde Award in 2018 from the IEEE Vehicular Technology Society, Joseph LoCicero Award in 2015 and Education Award in 2017 from the IEEE Communications Society, and Technical Recognition Award from Wireless Communications Technical Committee (2019) and AHSN Technical Committee (2013). Dr. Shen is the President of the IEEE Communications Society. He was the Vice President for Technical & Educational Activities, Vice President for Publications, Member-at-Large on the Board of Governors, Chair of the Distinguished Lecturer Selection Committee, Member of IEEE Fellow Selection Committee of the ComSoc. Dr. Shen served as the Editor-in-Chief of the IEEE IoT Journal, IEEE Network, and IET Communications.

"About the title" may belong to another edition of this title.

AHA-BUCH GmbH

Einbeck, Germany

5-star seller

AbeBooks seller since August 14, 2006

Shipping rates from Germany to U.S.A.

Item	5 to 7 business days	7 to 10 business days
First item	US$ 35.03	US$ 35.03

Delivery times are set by sellers and vary by carrier and location. Orders passing through Customs may face delays and buyers are responsible for any associated duties or fees. Sellers may contact you regarding additional charges to cover any increased costs to ship your items.

Payment methods

Bank Wire Transfer
Check
Paypal

Store description

Das Unternehmen AHA-BUCH GmbH: Seit der Gründung von AHA-BUCH im Juli 2005 ist unser Hauptziel, zufriedenen Kunden so schnell und so preisgünstig wie möglich ihren Bücherwunsch zu erfüllen. Unsere Firma beschäftigt 16 Mitarbeiter, die nur ein Ziel kennen: den Kunden und seine Wünsche! Auf über 3700 m2 Fläche haben wir über 100.000 Bücher, Modernes Antiquariat und Spiele auf Lager.

Specialty

Kinderbücher & Kinderhör Casetten, German Books, Software, Natur & Tiere, Ratgeber, Sachbücher, Englische Bücher, Medizin & Gesundheit, Universität & Studium

Seller's business information

AHA-BUCH GmbH

Garlebsen 48
Einbeck, Germany 37574

Contact sellerView this seller's itemsView seller's storefront

Terms of sale

General Terms and Conditions and Customer Information / Privacy Policy

I. General Terms and Conditions

§ 1 Basic provisions

(1) The following terms and conditions apply to all contracts that you conclude with us as a provider (AHA-BUCH GmbH) via the Internet platforms AbeBooks and/or ZVAB. Unless otherwise agreed, the inclusion of any of your own terms and conditions used by you will be objected to

(2) A consumer within the meaning of the following regulations is any natural person who concludes a legal transaction for purposes that can predominantly be attributed neither to their commercial nor their independent professional activity. An entrepreneur is any natural or legal person or a partnership with legal capacity who, when concluding a legal transaction, acts in the exercise of his independent professional or commercial activity.

§ 2 Conclusion of the contract

(1) The object of the contract is the sale of goods.

(2) If an article is listed by us at AbeBooks or ZVAB, the activation of the offer page at AbeBooks or ZVAB constitutes the binding offer to conclude a contract under the conditions contained in the article page.

(3) The contract is concluded via the online shopping cart system as follows:
The goods intended for purchase are stored in the "shopping cart". Via the corresponding button in the navigation bar you can call up the "shopping cart" and make changes there at any time.
After calling up the "Checkout" page and entering the personal data as well as the payment and shipping conditions, all order data will be displayed again on the order overview page.
Before sending the order, you have the option of checking all information again, changing it (also via the "back" function of the Internet browser) or cancelling the purchase.
By submitting the order via the "Buy now" button, you declare the acceptance of the offer in a legally binding manner, whereby the contract is concluded.

(4) The processing of the order and transmission of all information required in connection with the conclusion of the contract is partly automated by e-mail. You must therefore ensure that the e-mail address you have stored with us is correct, that the receipt of the e-mails is technically ensured and, in particular, that spam filters do not prevent it.

§ 3 Right of retention, retention of title

(1) You can only exercise a right of retention if it concerns claims from the same contractual relationship.

(2) The goods remain our property until full payment of the purchase price.

§ 4 Warranty

(1) The statutory warranty rights exist.

(2) In the case of used goods, the warranty period is one year from delivery of the item, deviating from the statutory regulation. The shortening of the deadline does not apply:

culpably caused damages attributable to us from injury to life, limb or health and in the case of other damages caused intentionally or through gross negligence;
insofar as we have fraudulently concealed the defect or have assumed a guarantee for the quality of the item.

(3) As a consumer, you are requested to check the item immediately upon delivery for completeness, obvious defects and transport damage and to inform us and the freight forwarder of any complaints as soon as possible. If you do not comply with this, this will have no effect on your statutory warranty claims.

§ 5 Choice of law, place of performance, place of jurisdiction

(1) German law shall apply. In the case of consumers, this choice of law shall only apply insofar as this does not withdraw the protection afforded by mandatory provisions of the law of the state of the consumer's habitual residence (principle of favourability).

(2) The place of performance for all services arising from the existing business relationships with us as well as the place of jurisdiction is our registered office, insofar as you are not a consumer, but a merchant, a legal entity under public law or a special fund under public law. The same applies if you do not have a general place of jurisdiction in Germany or the EU or if the domicile or habitual residence is not known at the time the action is brought. The right to appeal to the court at another statutory place of jurisdiction remains unaffected by this.

(3) The provisions of the UN Convention on Contracts for the International Sale of Goods shall expressly not apply.

II. Customer Information

Identity of the Seller

AHA-BUCH GmbH
Garlebsen 48
D-37574 Einbeck
Germany
Telephone: 055639996039
E-mail: abebooks@aha-buch.de.

Alternative dispute resolution:
The European Commission provides a platform for out-of-court online dispute resolution (ODR platform), which can be accessed under https://ec.europa.eu/odr.

Information on the conclusion of the contract

The technical steps for the conclusion of the contract, the conclusion of the contract itself and the correction options are carried out in accordance with the provisions "Conclusion of the contract" of our General Terms and Conditions (Part I.).

Contract language, contract text storage

3.1. The contract language is german.

3.2. The complete text of the contract will not be stored by us. Before sending the order, the contract data can be printed out or electronically saved via the print function of the browser. After receipt of the order by us, the order data, the legally prescribed information for distance contracts and the General Terms and Conditions will be sent to you again by e-mail.

Essential characteristics of the goods or services

The essential characteristics of the goods and/or services can be found in the respective offer.

Prices and terms of payment

5.1. The prices stated in the respective offers as well as the shipping costs represent total prices. They include all price components including all applicable taxes.

5.2. The shipping costs incurred are not included in the purchase price. They can be called up via a correspondingly designated button on our website or in the respective offer, are shown separately in the course of the ordering process and are to be borne by you in addition, unless free shipping has been promised.

5.3. If the delivery is made to countries outside the European Union, we may incur additional costs for which we are not responsible, such as.B customs duties, taxes or money transfer fees (transfer or exchange rate fees of the credit institutions), which are to be borne by you. Any costs incurred for the transfer of money shall also be borne by you in cases where the delivery is made to an EU member state, but the payment was initiated outside the European Union.

5.4. The payment methods available to you are indicated under a corresponding button on our website or in the respective offer.

5.5. Unless otherwise stated for the individual payment methods, the payment claims from the concluded contract are due for payment immediately.

Terms of delivery

6.1. The terms of delivery, the delivery date and, if applicable, existing delivery restrictions can be found under a correspondingly designated button on our website or in the respective offer.

Insofar as no other deadline is specified in the respective offer or under the correspondingly designated button, the delivery of the goods will take place within 3-5 days after conclusion of the contract (in the case of agreed advance payment, however, only after the time of your payment instruction).

6.2. Insofar as you are a consumer, it is regulated by law that the risk of accidental loss and accidental deterioration of the sold item during dispatch shall only pass to you upon handover of the goods, regardless of whether the shipment is insured or uninsured. This does not apply if you have independently commissioned a transport company not named by the entrepreneur or a person otherwise designated to carry out the shipment.

Statutory liability for defects

Liability for defects is governed by the "Warranty" provision in our General Terms and Conditions (Part I).

last update: 01/06/2022

Unless otherwise stated below, the provision of your personal data is neither required by law or contract, nor is it necessary for the conclusion of a contract. You are not obliged to provide the data. Failure to provide this will have no consequences. This only applies if no other information is provided in the subsequent processing operations.

Collection, processing and disclosure of personal data when placing orders
When ordering, we collect and process your personal data only to the extent necessary to fulfil and process your order and to process your enquiries. The provision of the data is necessary for the conclusion of the contract. Failure to provide this provision means that no contract can be concluded. The processing takes place on the basis of Art. 6 para. 1 lit.b GDPR and is necessary for the performance of a contract with you.
Your data will be passed on, for example, to the shipping companies and dropshipping providers selected by you, payment service providers, service providers for order processing and IT service providers.
In all cases, we strictly observe the legal requirements. The scope of data transmission is limited to a minimum.

Duration of storage
After completion of the contract, the data will first be stored for the duration of the warranty period, then taking into account legal, in particular tax and commercial retention periods and then deleted after expiry of the period, unless you have consented to further processing and use.

Rights of the data subject
If the legal requirements are met, you have the following rights under Articles 15 to 20 GDPR: Right to information, to correction, to erasure, to restriction of processing, to data portability.
In addition, pursuant to Article 21 (1) GDPR, you have the right to object to processing based on Article 6 (1) f GDPR and to processing for direct marketing purposes.

Right to lodge a complaint with the supervisory authority
In accordance with Article 77 GDPR, you have the right to complain to the supervisory authority if you believe that the processing of your personal data is not lawful.

Right to object
If the personal data processing listed here is based on our legitimate interest pursuant to Art. 6 para. 1 lit. f GDPR, you have the right to object to this processing at any time with effect for the future for reasons arising from your particular situation.

Shipping terms

We ship your order after we received them
for articles on hand latest 24 hours,
for articles with overnight supply latest 48 hours.
In case we need to order an article from our supplier our dispatch time depends on the reception date of the articles, but the articles will be shipped on the same day.

Our goal is to send the ordered articles in the fastest, but also most efficient and secure way to our customers.

Backdoor Attacks against Learning-Based Algorithms

Item description from seller

Bibliographic details

Synopsis

About the title

About the Author

AHA-BUCH GmbH

Shipping rates from Germany to U.S.A.

Payment methods

Store description

Specialty

Seller's business information

Terms of sale

Shipping terms