In another aspect, a system for machine learning using a distributed framework, includes a computing device communicatively connected to a plurality of remote devices, the computing device designed and configured to select at least a remote device of a plurality of remote devices, determine a confidence level of the at least a remote device, and assign at least a machine-learning task to the at least a remote device, wherein assigning further comprises assigning at least a secure data storage task to the at least a remote device and assigning at least a model-generation task to the at least a remote device.

A secret key value that is inaccessible to software is scrambled according to registers consisting of one-time programmable (OTP) bits. A first OTP register is used to change the scrambling of the secret key value whenever a lifecycle event occurs. A second OTP register is used to undo the change in the scrambling of the secret key. A third OTP register is used to affect a permanent change to the scrambling of the secret key. The scrambled values of the secret key (whether changed or unchanged) are used as seeds to produce keys for cryptographic operations by a device.

A secret key value that is inaccessible to software is scrambled according to registers consisting of one-time programmable (OTP) bits. A first OTP register is used to change the scrambling of the secret key value whenever a lifecycle event occurs. A second OTP register is used to undo the change in the scrambling of the secret key. A third OTP register is used to affect a permanent change to the scrambling of the secret key. The scrambled values of the secret key (whether changed or unchanged) are used as seeds to produce keys for cryptographic operations by a device.

A conversion device for secure computation for converting an input data which is an object data of secure computation into an input format applicable to the secure computation is provided. A conversion device for secure computation of the present invention includes an acquisition unit configured to acquire an object data of the secure computation; a storage unit configured to store a correspondence table specifying an input format required for executing the secure computation; a conversion processing unit configured to perform a conversion from the acquired object data into a secure computation data in accordance with the correspondence table; and an output unit configured to output the secure computation data.

An electronic watermark embedding apparatus according to an embodiment is an electronic watermark embedding apparatus capable of embedding an electronic watermark into a decoding circuit of secret-key encryption, and includes an embedding unit configured to generate the decoding circuit. The decoding circuit is embedded with the electronic watermark by being input with a common parameter generated in a setup of the secret-key encryption, a secret key of the secret-key encryption, and the electronic watermark, and is capable of decoding an encrypted text encrypted using the secret-key encryption.

A server determines an array [[addr]] indicating a storage destination of each piece of data, generates an array of concealed values, and connects the generated array to the array [[addr]] to determine an array [[addr′]]. The server generates a sort permutation [[σ.sub.1]] for the array, applies the sort permutation [[σ.sub.1]] to the array [[addr′]], and converts the array [[addr′]] into an array with a sequence composed of first Z elements set to [[i]] followed by α.sub.i elements set to [[B]]. The server generates a sort permutation [[σ.sub.2]] for the converted array [[addr′]], generates dummy data, imparts the generated dummy data to the concealed data sequence, applies the sort permutations [[σ.sub.1]] and [[σ.sub.2]] to the data array imparted with the dummy data, and generates, as a secret hash table, a data sequence obtained by deleting the last N pieces of data from the sorted data array.

An exemplary embodiment of the present disclosure provides a physically unclonable function (PUF) cell capable of exhibiting a stable performance and showing an excellent repeatability while being less affected by environmental factors such as a noise, temperature, and bias voltage. The PUF cell generates an output value by combining a scheme of amplifying a threshold voltage difference and a scheme of amplifying an oscillation frequency difference. In an oscillator that generates oscillation signals of different frequencies, the frequency difference of the oscillation signals is amplified by alternately supplying bias voltages of different magnitudes generated by utilizing the threshold voltage difference to a plurality of stages in the oscillator.

Secure analytics using homomorphic and injective format-preserving encryption are disclosed herein. An example method includes encoding an analytic parameter set using a homomorphic encryption scheme as a set of homomorphic analytic vectors; transmitting the set of homomorphic analytic vectors to a server system; and receiving a homomorphic encrypted result from the server system, the server system having utilized the homomorphic encryption scheme and a first injective, format-preserving encryption scheme to evaluate the set of homomorphic analytic vectors over a datasource.

Secure analytics using homomorphic and injective format-preserving encryption are disclosed herein. An example method includes encoding an analytic parameter set using a homomorphic encryption scheme as a set of homomorphic analytic vectors; transmitting the set of homomorphic analytic vectors to a server system; and receiving a homomorphic encrypted result from the server system, the server system having utilized the homomorphic encryption scheme and a first injective, format-preserving encryption scheme to evaluate the set of homomorphic analytic vectors over a datasource.

An aggregate sum is efficiently obtained while keeping confidentiality. A prefix-sum part computes a prefix-sum from a share of a sorted value attribute. A flag converting part converts a format of a share of a flag representing the last element of a group. A flag applying part generates a share of a vector in which a prefix-sum is set when a flag representing the last element of a group is true, and a sum of the whole is set when the flag is false. A sorting part generates a share of a sorted vector obtained by sorting a vector with a permutation which moves elements so that the last elements of each group are sequentially arranged from beginning. A sum computing part generates a share of a vector representing a sum for each group.