Diagnosing multi-stage diseases typically requires doctors to consider multiple data sources,including clinical symptoms,physical signs,biochemical test results,imaging findings,pathological examination data,and even ...Diagnosing multi-stage diseases typically requires doctors to consider multiple data sources,including clinical symptoms,physical signs,biochemical test results,imaging findings,pathological examination data,and even genetic data.When applying machine learning modeling to predict and diagnose multi-stage diseases,several challenges need to be addressed.Firstly,the model needs to handle multimodal data,as the data used by doctors for diagnosis includes image data,natural language data,and structured data.Secondly,privacy of patients’data needs to be protected,as these data contain the most sensitive and private information.Lastly,considering the practicality of the model,the computational requirements should not be too high.To address these challenges,this paper proposes a privacy-preserving federated deep learning diagnostic method for multi-stage diseases.This method improves the forward and backward propagation processes of deep neural network modeling algorithms and introduces a homomorphic encryption step to design a federated modeling algorithm without the need for an arbiter.It also utilizes dedicated integrated circuits to implement the hardware Paillier algorithm,providing accelerated support for homomorphic encryption in modeling.Finally,this paper designs and conducts experiments to evaluate the proposed solution.The experimental results show that in privacy-preserving federated deep learning diagnostic modeling,the method in this paper achieves the same modeling performance as ordinary modeling without privacy protection,and has higher modeling speed compared to similar algorithms.展开更多
This paper proposes a strategy for machine learning in the ciphertext domain.The data to be trained in the linear regression equation is encrypted by SHE homomorphic encryption,and then trained in the ciphertext domai...This paper proposes a strategy for machine learning in the ciphertext domain.The data to be trained in the linear regression equation is encrypted by SHE homomorphic encryption,and then trained in the ciphertext domain.At the same time,it is guaranteed that the error of the training results between the ciphertext domain and the plaintext domain is in a controllable range.After the training,the ciphertext can be decrypted and restored to the original plaintext training data.展开更多
In this paper,we provide a new approach to data encryption using generalized inverses.Encryption is based on the implementation of weighted Moore–Penrose inverse A y MNenxmT over the nx8 constant matrix.The square He...In this paper,we provide a new approach to data encryption using generalized inverses.Encryption is based on the implementation of weighted Moore–Penrose inverse A y MNenxmT over the nx8 constant matrix.The square Hermitian positive definite matrix N8x8 p is the key.The proposed solution represents a very strong key since the number of different variants of positive definite matrices of order 8 is huge.We have provided NIST(National Institute of Standards and Technology)quality assurance tests for a random generated Hermitian matrix(a total of 10 different tests and additional analysis with approximate entropy and random digression).In the additional testing of the quality of the random matrix generated,we can conclude that the results of our analysis satisfy the defined strict requirements.This proposed MP encryption method can be applied effectively in the encryption and decryption of images in multi-party communications.In the experimental part of this paper,we give a comparison of encryption methods between machine learning methods.Machine learning algorithms could be compared by achieved results of classification concentrating on classes.In a comparative analysis,we give results of classifying of advanced encryption standard(AES)algorithm and proposed encryption method based on Moore–Penrose inverse.展开更多
基金funded by the National Natural Science Foundation,China(No.62172123)the Key Research and Development Program of Heilongjiang(Grant No.2022ZX01A36)+1 种基金the Special Projects for the Central Government to Guide the Development of Local Science and Technology,China(No.ZY20B11)the Harbin Manufacturing Technology Innovation Talent Project(No.CXRC20221104236).
文摘Diagnosing multi-stage diseases typically requires doctors to consider multiple data sources,including clinical symptoms,physical signs,biochemical test results,imaging findings,pathological examination data,and even genetic data.When applying machine learning modeling to predict and diagnose multi-stage diseases,several challenges need to be addressed.Firstly,the model needs to handle multimodal data,as the data used by doctors for diagnosis includes image data,natural language data,and structured data.Secondly,privacy of patients’data needs to be protected,as these data contain the most sensitive and private information.Lastly,considering the practicality of the model,the computational requirements should not be too high.To address these challenges,this paper proposes a privacy-preserving federated deep learning diagnostic method for multi-stage diseases.This method improves the forward and backward propagation processes of deep neural network modeling algorithms and introduces a homomorphic encryption step to design a federated modeling algorithm without the need for an arbiter.It also utilizes dedicated integrated circuits to implement the hardware Paillier algorithm,providing accelerated support for homomorphic encryption in modeling.Finally,this paper designs and conducts experiments to evaluate the proposed solution.The experimental results show that in privacy-preserving federated deep learning diagnostic modeling,the method in this paper achieves the same modeling performance as ordinary modeling without privacy protection,and has higher modeling speed compared to similar algorithms.
文摘This paper proposes a strategy for machine learning in the ciphertext domain.The data to be trained in the linear regression equation is encrypted by SHE homomorphic encryption,and then trained in the ciphertext domain.At the same time,it is guaranteed that the error of the training results between the ciphertext domain and the plaintext domain is in a controllable range.After the training,the ciphertext can be decrypted and restored to the original plaintext training data.
基金the support of Network Communication Technology(NCT)Research Groups,FTSM,UKM in providing facilities for this research.This paper is supported under the Dana Impak Perdana UKM DIP-2018-040 and Fundamental Research Grant Scheme FRGS/1/2018/TK04/UKM/02/7.
文摘In this paper,we provide a new approach to data encryption using generalized inverses.Encryption is based on the implementation of weighted Moore–Penrose inverse A y MNenxmT over the nx8 constant matrix.The square Hermitian positive definite matrix N8x8 p is the key.The proposed solution represents a very strong key since the number of different variants of positive definite matrices of order 8 is huge.We have provided NIST(National Institute of Standards and Technology)quality assurance tests for a random generated Hermitian matrix(a total of 10 different tests and additional analysis with approximate entropy and random digression).In the additional testing of the quality of the random matrix generated,we can conclude that the results of our analysis satisfy the defined strict requirements.This proposed MP encryption method can be applied effectively in the encryption and decryption of images in multi-party communications.In the experimental part of this paper,we give a comparison of encryption methods between machine learning methods.Machine learning algorithms could be compared by achieved results of classification concentrating on classes.In a comparative analysis,we give results of classifying of advanced encryption standard(AES)algorithm and proposed encryption method based on Moore–Penrose inverse.
