103results about "Biomolecular computers" patented technology

Systems and methods described herein include embodiments for generating a geometrical structure of a binding complex formed between a plurality of macromolecules, comprising: processing an input representation comprising a plurality of representations of the plurality of macromolecules to generate a geometry prior; sampling an initial geometrical structure of the binding complex based on the geometry prior; and processing, using a neural network, the initial geometrical structure to generate the geometrical structure of the binding complex formed by the plurality of macromolecules.

The invention discloses an adaptive distributed network threat detection and response system and method, and the system comprises a data collection module which is used for capturing network traffic in real time and converting the network traffic into structured data; the data analysis module is used for performing parallel analysis on the structured data by adopting a plurality of anomaly detection algorithms; the algorithm fusion module is used for counting the abnormal judgment times of each algorithm on the same sample through a voting mechanism, and outputting a final classification result according to a preset threshold value; the method comprises the following working steps: concurrently capturing network traffic through distributed nodes, analyzing the network traffic into structured data, operating an isolated forest algorithm, a DBSCAN algorithm and a K-means algorithm in parallel, and outputting an anomaly judgment result; and fusing multi-algorithm results and generating final classification through a voting mechanism. The invention provides an efficient, intelligent and distributed network security protection system, which can effectively detect and respond to various network threats, and is particularly suitable for large-scale distributed environments and dynamic network security defense requirements.

The invention discloses a traffic control and guidance system and method based on an ant colony algorithm, belongs to the technical field of road traffic control, and solves the problems that a neural network model in an existing method mainly focuses on respective route conditions of a plurality of bifurcation routes of a bifurcation on a one-way driving road, and the information fusion degree between traffic elements is low. The method comprises the following steps: identifying dynamic characteristic information of vehicles in a road network, pre-constructing a road network optimization model based on an ant colony algorithm combined with deep learning, and analyzing a pheromone spread function cluster solution; according to the invention, the road network optimization model based on the ant colony algorithm and the deep learning is pre-constructed, the collaborative optimization of traffic signal control and vehicle induction is realized, the traffic signal optimization control strategy and the vehicle optimization path strategy are output, so that the signal timing and the induction path can be matched with each other, and the control accuracy is improved. Therefore, the operation efficiency of the traffic system is improved and the accuracy and effectiveness of the guidance strategy are ensured.

A computing architecture is optimized for the reuse of machine learning model. A request is received for a machine learning model recommendation that specifies parameters associated with a desired machine learning model (e.g., schema, etc.). Thereafter, a machine learning-based discovery model, recommends at least one machine learning model (to reuse) based on the parameters specified in the request. Next, data characterizing the recommended at least one machine learning model is provided (e.g., loaded into memory, displayed in a graphical user interface, transmitted to a remote computing system, and / or stored in physical persistence, etc.). In some variations, a graphical user interface can be rendered that allows a client application to select one of the recommended machine learning models and, further optionally, to activate the selected machine learning models. Related apparatus, systems, techniques and articles are also described.

An artificial reality system performs local user adaptation of machine learning models and global improvement of the machine learning models while ensuring data security and privacy. An artificial reality headset uses smart sensors and a system on chip (SoC). The smart sensor extracts features from sensor data, encrypts them and transmits the encrypted features to the SoC of the headset. The SoC provides the encrypted features to a server, for example, a server in a cloud platform. The server updates a base model with collected features and sends the updated base model parameters to headsets. The smart sensors of a headset receiving the updated model parameters, balance the model parameters with locally stored feature data, for example, user calibration features to perform user customization of the machine learning model.

The invention relates to the technical field of electromagnetic metasurfaces and bionic computing, and discloses a bionic information processing method and system based on an electromagnetic metasurface. The method comprises the following steps: S1, mapping DNA sequence information into a multi-dimensional coding state of the bionic metasurface information processing unit in an electromagnetic adjustable parameter space; s2, realizing co-evolution of the multi-dimensional coding state in space, frequency, phase and amplitude dimensions by dynamically regulating and controlling electromagnetic response parameters of the bionic metasurface information processing unit; and S3, using propagation and interference of electromagnetic waves in the diffraction neural network and a near-field coupling effect between the bionic metasurface information processing units to complete copying, logical operation and information read-write operation of the DNA sequence information. The functions of DNA coding, copying, logic calculation, information reading and writing and the like are uniformly mapped into the programmable electromagnetic metasurface platform, and cooperative processing of information in the aspects of space, electromagnetism and algorithm is achieved.

Aspects herein describe new methods of providing explainable solutions to domain-specific problems using artificial intelligence. A domain-specific problem, for example, a biological problem, is identified by a computational unit. The computational unit determines an initial feature subset comprising various features for observable characteristics of the problem. The computational unit optimizes the initial feature subset to improve the accuracy of the subset in generating a solution to the problem. The computational unit trains one or more predictive artificial intelligence models, based on the optimized subset, to output solutions to the problem. Subsequently, the computational unit utilizes the trained predictive model to output solutions to one or more additional domain-specific problems. The computational unit may additionally provide explainable solutions to the domain-specific problems by outputting a representation of steps performed by the predictive model to output the solutions.

The invention discloses an industrial control network risk monitoring method based on immune recognition, and relates to the technical field of network security risk monitoring, and the method comprises the steps: obtaining multi-source heterogeneous industrial control protocol flow data and a known data set which are collected in real time; preprocessing the data to obtain a standardized antigen vector; according to the standardized antigen vector, carrying out mixed training on a preset depth auto-encoder and a graph neural network model based on an immune algorithm to obtain a trained depth auto-encoder-graph neural network model and an initial detector population; inputting the standardized antigen vector into a lightweight model to obtain a compressed antigen vector; and detecting the compressed antigen vector based on the initial detector population to obtain an antigen vector for triggering an alarm. According to the method, the definition and expandability of original protocol semantics are improved, continuous coverage of unknown attacks is ensured, and the deployment and updating process of the model is optimized.

A biological computing platform may include a multielectrode array (MEA) connected to a computing device. The MEA may include a 2D grid of excitation sites, biological neurons disposed on the MEA, a processing device to apply a plurality of impulses at excitation sites having coordinates on the 2D grid, and one or more sensors to measure electrical signals output by one or more of the biological neurons at coordinates of the 2D grid, wherein the processing device is to receive the electrical signals from the one or more sensors and generate a representation of the electrical signals. The computing device may be programmed to receive a digital input signal, convert the digital input signal into instructions for the plurality of impulses, send the instructions to the MEA, receive the representation of the electrical signals from the MEA, and process the representation of the electrical signals.

The governed interaction interface maintains an identity-preserving semantic-state for a knowledge entity and updates it only through validated semantic evidence generated autonomously by execution mechanisms. Semantic intentions describe semantic needs and are transformed into capability-requests containing no operational commands. Execution mechanisms—including environmental devices, perception models, symbolic analyzers, virtual actors, and software agents—interpret capability-requests independently through tool-side semantic interpreters and may act or decline to act. The system does not observe or evaluate tool behavior and receives only semantic evidence describing semantic meaning of any resulting effect. The governance engine evaluates each evidence fragment independently under identity, coherence, lineage, evidentiary sufficiency, deviation constraints, and contextual compatibility. Semantic-state transitions occur only when validated evidence satisfies governance constraints. The system performs no prediction, optimization, control computation, multimodal fusion, or supervisory coordination, remaining fully separated from device-level behavior.

A Biological Neural Network (BNN) core unit comprising a neural cell culture, an input stimulation unit, an output readout unit may be controlled through its various life cycles to provide data processing functionality. An automation system comprising an environmental and chemical controller unit adapted to operate with the BNN stimulation and readout data interfaces facilitates the monitoring and adaptation of the BNN core unit parameters. Pre-processing and post-processing of the BNN interface signals may further facilitate the training and reinforcement learning by the BNN. Multiple BNN core units may also be assembled together as a stack. The proposed system provides a BNN Operating System as a core component for a wetware server to receive, process and transmit data for different client applications without exposing the BNN core unit components to the client user while requiring significantly less energy than conventional silicon-based hardware and software information processing for high-level cognitive computing tasks.

A Biological Neural Network (BNN) core unit comprising a neural cell culture, an input stimulation unit, an output readout unit may be controlled through its various life cycles to provide data processing functionality. An automation system comprising an environmental and chemical controller unit adapted to operate with the BNN stimulation and readout data interfaces facilitates the monitoring and adaptation of the BNN core unit parameters. Pre-processing and post-processing of the BNN interface signals may further facilitate the training and reinforcement learning by the BNN. Multiple BNN core units may also be assembled together as a stack. The proposed system provides a BNN Operating System as a core component for a wetware server to receive, process and transmit data for different client applications without exposing the BNN core unit components to the client user while requiring significantly less energy than conventional silicon-based hardware and software information processing for high-level cognitive computing tasks.

The invention discloses a cellular automaton-based land utilization and urban element collaborative simulation method. The method comprises the following steps of: calculating the development probability of each element; the method comprises the following steps of: constructing a cellular automata (CA) model of each element; space collaborative simulation calculation; and evaluating the result precision. According to the method, the urban land use change simulation precision can be effectively improved, the land use change in reality can be better simulated, the simulation precision of population and economic elements can be improved, and the interaction and synergistic effect between different urban elements are effectively revealed. By means of multi-element collaborative simulation, small changes of urban development elements can be caught more finely, and the method has good perception ability for residential land, industrial land, commercial land, public management land and the like and has important significance for simulating urban collaborative development and reasonably planning urban layout.

The invention discloses an energy storage capacity regulation and control method and system of a power generation system based on complementation of light, wind and electricity, relates to the field of energy storage capacity regulation and control, and ensures that the energy storage capacity influence parameters of various types and energy storage capacity regulation and control strategies are not influenced in the power generation process of the current power generation system by constructing a mapping equation between the energy storage capacity influence parameters and the energy storage capacity regulation and control strategies. Real-time energy storage capacity influence parameters are acquired at intervals, so that a corresponding optimal energy storage capacity regulation and control strategy is mapped; besides, the time interval for acquiring the real-time energy storage capacity influence parameter is adjusted by acquiring the percentage data of the time period in which the power generation capacity of the current power generation system in a period cannot meet the electric quantity required by the electric load in the current energy storage capacity regulation and control strategy adjustment evaluation period; and it is ensured that the percentage data of the time period of the electric quantity required by the electric load in the current energy storage capacity regulation and control strategy adjustment evaluation period meets the requirement.

Disclosed are metalloproteins (e.g, vanadyl-substituted metalloproteins), such as metal-substituted rubredoxins including vanadyl-substituted rubredoxins. These metalloproteins can be used in quantum computing applications, for example, as quantum bits (qubits) or quantum dits (qudits).

The invention provides an intelligent terminal outdoor cooperation method and system based on an ant colony algorithm. The method comprises the following steps: acquiring outdoor navigation data, processing the outdoor navigation data into navigation data fragments, and distributing the navigation data fragments to a plurality of intelligent terminals; screening out a main relay node; when entering the target area, determining the intelligent terminal as main navigation equipment, and controlling the gas intelligent terminal to enter a low-power-consumption mode; summarizing the navigation data fragments stored in the intelligent terminals through the main relay node, forwarding the summarized navigation data fragments to the main navigation equipment, and splicing to generate complete navigation data of the target area; and the main navigation equipment obtains and generates navigation information and synchronizes the navigation information to each intelligent terminal through the main relay node. Through pheromone dynamic iteration of the ant colony algorithm and a distributed decision-making mechanism, efficient cooperation of multiple intelligent terminals on division storage and group collaborative splicing is realized, and on the premise of guaranteeing complete navigation data splicing precision and real-time navigation reliability, the overall energy consumption is remarkably reduced, and the equipment endurance is prolonged.

Embodiments of the invention describe systems (110) and methods (400) for monitoring ripening of perishable items (104) stored within a container (102). The method (400) comprises measuring (402) one or more environmental parameters within the container (102) at pre-determined time intervals. Further, the method (400) comprises determining (404) a respiration rate associated with each of the perishable items (104) based on the measured one or more environmental parameters. Further, the method (400) comprises determining (406), based on the determined respiration rate and a ripening database (204a), an initiation time for initiating ripening of the perishable items (104). Furthermore, the method (400) comprises triggering (408) the ripening of the perishable items (104) at the initiation time.

Systems and techniques are provided for making genetically engineered ion channels (ICs) with bioluminescent resonance energy transfer (BRET) complexes and using such ICs for efficient readout of neural activity and output of biological neuronal networks.SOLUTION: In one implementation, the disclosed technology includes identifying a target location in the IC for expression of a target protein including a donor tag protein and an acceptor tag protein, and modifying the genome of the neuronal cell at a portion associated with the target location in the IC. The technique further includes causing the neuronal cell to express the target protein in the IC according to the modified genome. In the first (second) state of the IC, the donor tag protein is at a first (second) distance from the acceptor tag protein that is related to the absence (presence) of energy transfer between the donor tag protein and the acceptor tag protein.SELECTED DRAWING: Figure 3A-3C

Embodiments of the present disclosure describe systems and methods for monitoring the maturation of perishable items stored within a container. The method includes measuring one or more environmental parameters within the container at predetermined time intervals. Further, the method includes determining a respiration rate associated with each of the perishable items based on the measured one or more environmental parameters. Further, the method includes determining an initiation time for initiating aging of the perishable item based on the determined respiration rate and an aging database. Further, the method includes triggering curing of the perishable article at the start-up time.

A nanosynapse integrated device and a manufacturing method therefor. The nanosynapse integrated device comprises: a substrate (1); an insulation layer (2) provided on the substrate (1); and a plurality of nanosynapse units (3) provided on the insulation layer (2). Each nanosynapse unit (3) comprises a first microelectrode (31) and a second microelectrode (32) which are opposite, and a GaN-based nanowire (33) provided between the first microelectrode (31) and the second microelectrode (32), a first end of the GaN-based nanowire (33) being bonded to the first microelectrode (31), and a second end of the GaN-based nanowire (33) being bonded to the second microelectrode (32). The nanosynapse integrated device uses the GaN-based nanowires (33) as synapse elements in the device, and the GaN-based nanowires (33) have controllable direct band gaps and stable physical and chemical properties. In the nanosynapse integrated device, the plurality of nanosynapse units (3) are integrally arranged on the same substrate (1), and therefore the nanosynapse integrated device can be manufactured on a scale, and has a small size and high level of integration. The plurality of nanosynapse units (3) can be separately and independently controlled or can be subjected to overall array control according to an arranged array pattern.

The disclosed technology relates to constructing a convolutional neural network based classifier for variant classification. More specifically, it relates to training a convolutional neural network based classifier on training data using a back-propagation gradient update method, which gradually matches the output of the convolutional neural network based classifier with the corresponding true value label. The convolutional neural network based classifier includes a residual block group, each residual block group is parameterized by the number of convolution filters in the residual block, the convolution window size of the residual block, and the hole convolution rate of the residual block, the convolution window size varies between the residual block groups, and the hole convolution rate varies between the residual block groups. The training data includes benign training examples and pathogenic training examples of translation sequence pairs generated from benign variants and pathogenic variants.

Aspects herein describe a new approach to providing an interpretable solution for a domain-specific problem using artificial intelligence. Problems of a particular domain, such as biological problems, are identified by a computing unit. The computing unit determines an initial subset of features including various features for observable characteristics of the problem. The computing unit optimizes the subset of initial features to improve the accuracy of the subset in generating a solution to the problem. The computing unit trains one or more predictive artificial intelligence models based on the optimized subset to output a solution to the problem. The computing unit then uses the trained predictive model to output solutions to the one or more additional domain-specific problems. The computing unit may additionally provide an interpretable solution for a problem of a particular field by outputting a representation of steps performed by the predictive model to output the solution.

Techniques for identifying types of cells present in a biological sample using flow cytometry performed using a panel of markers and multiple machine learning models. The techniques include: obtaining flow cytometry data for the biological sample, the biological sample previously-obtained from a subject and comprising a plurality of cells, the flow cytometry data including flow cytometry measurements obtained during respective flow cytometry events, the flow cytometry events including a subset of events corresponding to cells in the biological sample being measured by the flow cytometry platform; and identifying types of cells of the plurality of cells using the multiple machine learning models to obtain a respective plurality of cell types.

The invention discloses an emotion memory generation system based on a memristor, and belongs to the field of integrated circuit design. The system comprises a brain emotion learning model and a hippocampus model. The brain emotion learning model comprises a thalamus module, a sensory cortex module, an orbital frontal cortex module and an amygdala kernel module; the hippocampus model comprises a DG module, a CA3 module and a CA1 module; under the stimulation of the emotional signals, the thalamus module firstly preprocesses the emotional signals, then the sensory cortex module and the orbital frontal lobe module integrate and process the preprocessed emotional signals, and the processed signals are respectively transmitted to the CA3 module and the amygdala kernel module to generate output signals representing emotions; under the stimulation of a learning signal, the DG module firstly preprocesses the learning signal, then the CA3 region is responsible for receiving the learning signal and generating an output signal representing short-term memory, and the output signal is transmitted to the CA1 region; and after receiving the short-term memory signal, the CA1 area converts short-term memory into long-term memory and stores the long-term memory.

The invention discloses a carbon dioxide-starch biosynthesis intelligent regulation and control chip system based on multi-mode AI. The system comprises a photonic crystal biosensor array used for monitoring physical and chemical parameters such as a pH value, a dissolved oxygen concentration and a carbon dioxide concentration in a reaction process in real time; the 16 * NPU array is used for efficiently calculating and processing the data of the multi-modal sensor and predicting the reaction dynamic state in real time; the 4 * Cortex-M7 treatment core is used for controlling the reaction process in real time and adjusting reactor conditions; and a neural symbol system-based multi-modal reaction regulation and control algorithm which is used for optimizing reaction dynamics and realizing cooperative regulation of 2000 + reaction parameters. All parameters in the reaction process are optimized through real-time multi-mode data monitoring (including enzyme activity, reaction temperature, pH value and the like). The enzyme activity can be accurately adjusted, so that the purity of the final product is remarkably improved, and the purity of the starch can be stabilized at 99% or above, while the purity of the starch prepared by the traditional method is only 89.7%.

The invention belongs to the technical field of vehicle insurance assessment, and discloses a vehicle operation risk assessment method and device, computer equipment and a storage medium. An accident record of a to-be-evaluated vehicle and driving behavior characteristic data of a driver are acquired, and the driving behavior characteristic data comprise bad driving behavior frequency, night driving duration, fatigue driving duration, load data, speed data and dangerous driving behavior frequency; determining a plurality of risk evaluation parameters based on an accident record and the driving behavior characteristic data; determining the risk level of each risk evaluation parameter according to a preset risk level division standard; and inputting the risk evaluation parameters and the corresponding risk levels into a risk evaluation model to obtain an estimated operation risk value of the to-be-evaluated vehicle, the risk evaluation model being constructed based on a dendritic cell algorithm. And the operation risk of the heavy-duty truck can be estimated more accurately.

To provide a circuit and a method for computing a neural network inference in hardware.SOLUTION: An architecture 300 including a matrix computation unit, includes a two-dimensional systolic array 306 including a plurality of cells, and a weight fetcher interface. For each of a plurality of neural network layers, the weight fetcher interface sends a plurality of weight inputs to cells along a first dimension of the two-dimensional systolic array for the neural network layer. For each of the plurality of neural network layers, a plurality of weight sequencers, each of which is coupled to a distinct cell along the first dimension of the two-dimensional systolic array, shift, for the neural network layer, the plurality of weight inputs to cells along the second dimension of the two-dimensional systolic array over a plurality of clock cycles. The cells in the two-dimensional systolic array each compute a product of an activation input and a respective weight input using a multiplication circuit.SELECTED DRAWING: Figure 3

Provided herein are novel Class 2 Type II and Type V CRISPR-Cas RNA-guided endonucleases such as Cas9 and Cas12 endonucleases and systems comprising the same. Methods of making and methods of using the same are also provided. Exemplary methods of use include modifying target DNA and detecting targeted DNA, which can be useful for therapeutic and diagnostic applications. Some diagnostic applications can take advantage of the collateral nuclease activity of the enzyme bound to the target sequence.

The present invention is directed to a single input multiplex decision expression system, and a method of using same, such as for expressing a reporter protein.

To disclose a technique relating to constructing a convolutional neural network-based classifier for variant classification.SOLUTION: A disclosed technique relates to training a convolutional neural network-based classifier regarding training data by using a backpropagation-based gradient update technique that progressively collates an output of a convolutional neural network-based classifier with a corresponding ground truth level. The convolutional neural network-based classifier includes groups of residual blocks. Each group of residual blocks is parameterized by the number of convolution filters in the residual blocks, a convolution window size of the residual blocks, and an expansion convolution rate of the residual blocks, the convolution window size varies between the groups of the residual blocks, and the expansion convolution rate varies between the groups of the residual blocks. Training data includes a benign training example and a pathogenic training example of a translated sequence pair generated from a benign variant and a pathogenic variant.SELECTED DRAWING: Figure 19