A method for key agreement between a first party and a second party over a public communications channel, the method including selecting, by the first party, a first value “a”; multiplying the first value “a” by a second value “b” using Knuth multiplication to create a third value “d”, the third value “d” being a semistandard tableau; sending the third value “d” to the second party; receiving, from the second party, a fourth value “e”, the fourth value being a second semistandard tableau comprising the second value “b” multiplied by a fifth value “c” selected by the second party; and creating a shared secret by multiplying the first value “a” with the fourth value “e” using Knuth multiplication, wherein the shared secret matches the third value “d” multiplied by the fifth value “c” using Knuth multiplication.

A method for protecting a machine learning model is provided. In the method, a first machine learning model is trained, and a plurality of machine learning models derived from the first machine learning model is trained. Each of the plurality of machine learning models may be different from the first machine learning model. During inference operation, a first input sample is provided to the first machine learning model and to each of the plurality of machine learning models. The first machine learning model generates a first output and the plurality of machine learning models generates a plurality of second outputs. The plurality of second outputs are aggregated to determine a final output. The final output and the first output are classified to determine if the first input sample is an adversarial input. If it is adversarial input, a randomly generated output is provided instead of the first output.

A vehicle includes: at least one memory configured to store at least one default Instruction Structure Key (ISK), a generated ISK, and a pin code of the vehicle; and at least one processor. The at least one default ISK may include a first default ISK and a second default ISK. The processor may generate a random number using the first default ISK, receive the second default ISK encrypted with the generated ISK generated based on the pin code, and determine the generated ISK as an encryption key for encryption communication of the vehicle when the generated random number and the random number corresponding to the second default ISK are the same.

A power supplier circuit supplies a power signal to a codec of an audio apparatus. The power supplier circuit includes a random sequence generation circuit, a control circuit, and a power circuit. The random sequence generation circuit generates a random sequence. The control circuit outputs a first control signal according to the random sequence, a first reference signal, and the power signal. The power circuit generates the power signal according to the first control signal, such that the power signal is spread in response to the random sequence.

In one aspect, an electronic gaming system is described. The electronic gaming system includes a display device, a random number generator (RNG), a player input interface, and a game controller configured to execute instructions stored in a tangible, non-transitory, computer-readable medium. When executed by the game controller, the instructions cause the game controller to at least determine, during play of a base game, that an enhancement trigger condition is satisfied and convert the symbol displayed on each position having one of the first frame and the second frame displayed thereon into a wild symbol. The instructions further cause the game controller to determine that a bonus game trigger condition is satisfied and convert a plurality of positions on each reel with at least one position having the second frame displayed thereon, into a bonus reel. The instructions also cause the game controller to generate a bonus game outcome.

A data collection and analysis method includes applying a first noise step to an original data stream with an original character to generate a first data stream with a first character; and applying a second noise step to the first data stream to generate a second data stream with a second character, wherein a first variation between the original character and the first character is greater than a second variation between the original character and the second character.

Embodiments of the present disclosure provide a random number generator circuit, including: a random number generator, configured to output a plurality of first random numbers in each counting cycle; a control signal generation module, configured to receive a trigger signal and output control signals corresponding to different first random numbers based on the trigger signal; and a multi-select module, configured to receive the first random number and the control signal corresponding to the first random number, based on the control signal to adjust at least one bit position of the first random number, obtain a second random number, and output a plurality of the second random numbers.

A test stimulus generator generates error irritations, or error sequences, within a processor system. The test stimulus generator includes an initialization register, a pseudorandom number generator (PRNG), a clock subsystem, and an output register. The PRNG calculates an output value from an initialization value stored in the initialization register. The PRNG output value represents a unique error irritation and identifies one or more components within the processor system to handle the error irritation. The clock subsystem generates either a continuous or pulsed clock signal that transfers the initialization value into the PRNG. The output register stores the PRNG output value and transmits the corresponding error irritation to the processor components identified to handle the error irritation. The test stimulus generator generates error irritations in a predetermined or random order based on the initialization value. A corresponding method and computer program product are also disclosed.

A method for training an artificial neural network, ANN, which comprises a multiplicity of processing units. Parameters that characterize the behavior of the ANN are optimized according to a cost function. Depending on outputs determined from learning input quantity values and on learning output quantity values, an output of at least one selected processing unit is deactivated. Selection of the selected processing unit is achieved with the aid of a sequence of quasi-random numbers.

Described herein are systems and techniques for privacy-preserving unsupervised learning. The disclosed system and methods can enable separate computers, operated by separate entities, to perform unsupervised learning jointly based on a pool of their respective data, while preserving privacy. The system improves efficiency and scalability, while preserving privacy and avoids leaking a cluster identification. The system can jointly compute a secure distance via privacy-preserving multiplication of respective data values x and y from the computers based on a 1-out-of-N oblivious transfer (OT). In various embodiments, N may be 2, 4, or some other number of shares. A first computer can express its data value x in base-N. A second computer can form an ×N matrix comprising random numbers m.sub.i,0 and the remaining elements m.sub.i,j=(yjN.sup.i-m.sub.i,0) mod . The first computer can receive an output vector from the OT, having components m.sub.i=(yx.sub.i N.sup.i-m.sub.i,0) mod .