Abstract: A rapid mining method for sandstone-type uranium resources in a uranium and coal superposed area, which relates to the technical field of mining engineering is provided. The method includes: arranging a high-density adjustable well pattern in an in-situ leaching mining area; determining a length and a position of a filter located on the in-situ leaching mining area through a digital well construction technology; and in a production stage, carrying out operations such as pumping/injection centralized filtration, intensified leaching, high-intensity extraction, high-intensity injection, and change of layout of the high-density adjustable well pattern to rapidly obtain sandstone-type uranium resources in a uranium and coal superposed area. A recovery speed of the sandstone-type uranium resources can be improved, and service life of the in-situ leaching mining area can be shortened.

Abstract: Provided are a pharmaceutical composition of (S)-ketorolac and a preparation method therefor, which belong to the technical field of pharmaceutical preparations. The pharmaceutical composition includes (S)-ketorolac or a pharmaceutically acceptable salt thereof, a stabilizer, a pH-adjusting agent, and an excipient, and the dosage form of the pharmaceutical composition is preferably lyophilized powder injection. The lyophilized powder injection prepared herein has optical purity of an active ingredient greater than or equal to 95% after being stored under long-term stability test conditions for 6 months and optical purity of an active ingredient greater than or equal to 90% after being stored under accelerated stability test conditions for 6 months. The lyophilized powder injection has important use in the preparation of non-steroidal anti-inflammatory drugs with analgesic, anti-inflammatory, antipyretic effects.

Abstract: Generally, the present disclosure is directed to user interface understanding. More particularly, the present disclosure relates to training and utilization of machine-learned models for user interface prediction and/or generation. A machine-learned interface Nprediction model can be pre-trained using a variety of pre-training tasks for eventual downstream task training and utilization (e.g., interface prediction, interface generation, etc.).

Abstract: Techniques for using enhanced logit values for classifying utterances and messages input to chatbot systems in natural language processing. A method can include a chatbot system receiving an utterance generated by a user interacting with the chatbot system and inputting the utterance into a machine-learning model including a series of network layers. A final network layer of the series of network layers can include a logit function. The machine-learning model can map a first probability for a resolvable class to a first logit value using the logit function. The machine-learning model can map a second probability for a unresolvable class to an enhanced logit value. The method can also include the chatbot system classifying the utterance as the resolvable class or the unresolvable class based on the first logit value and the enhanced logit value.

Abstract: Techniques for using logit values for classifying utterances and messages input to chatbot systems in natural language processing. A method can include a chatbot system receiving an utterance generated by a user interacting with the chatbot system. The chatbot system can input the utterance into a machine-learning model including a set of binary classifiers. Each binary classifier of the set of binary classifiers can be associated with a modified logit function. The method can also include the machine-learning model using the modified logit function to generate a set of distance-based logit values for the utterance. The method can also include the machine-learning model applying an enhanced activation function to the set of distance-based logit values to generate a predicted output. The method can also include the chatbot system classifying, based on the predicted output, the utterance as being associated with the particular class.

Abstract: Techniques are disclosed herein for objective function optimization in target based hyperparameter tuning. In one aspect, a computer-implemented method is provided that includes initializing a machine learning algorithm with a set of hyperparameter values and obtaining a hyperparameter objective function that comprises a domain score for each domain that is calculated based on a number of instances within an evaluation dataset that are correctly or incorrectly predicted by the machine learning algorithm during a given trial. For each trial of a hyperparameter tuning process: training the machine learning algorithm to generate a machine learning model, running the machine learning model in different domains using the set of hyperparameter values, evaluating the machine learning model for each domain, and once the machine learning model has reached convergence, outputting at least one machine learning model.

Abstract: Methods are provided for improving performance of a computing system used to model potential crop yield. In one example embodiment, a computer-implemented method includes generating a model of potential crop yield, as a function of planting date and relative maturity based, at least in part, on one or more relative maturity maps, one or more planting date maps, and one or more actual production history maps, and storing the model in a memory of the server computer system. The method also includes receiving, via an interface at a field manager computing device, a selection of a particular field and computing, from the model of potential crop yield, a potential yield for the particular field based, at least in part, on a planting date for the particular field, a relative maturity value, and values representing actual production history for the particular field.

Abstract: A computer-implemented technique for decomposing semi-structured data is provided. In this technique, metadata for a predetermined number of records can be collected from semi-structured data that includes several records. A structured format is generated based on the metadata and the plurality of records is decomposed with the structured format.

Abstract: Techniques are disclosed herein for continuous hyperparameter tuning with automatic domain weight adjustment based on periodic performance checkpoints. In one aspect, a method is provided that includes initializing a machine learning algorithm with a set of hyperparameter values and obtaining a hyperparameter objective function that is defined at least in part on a plurality of domains of a search space that is associated with the machine learning algorithm. For each trial of a hyperparameter tuning process: running the machine learning algorithm in different domains using the set of hyperparameter values, periodically checking a performance of the machine learning algorithm in the different domains based on the hyperparameter objective function; and continuing hyperparameter tuning with a new set of hyperparameter values after automatically adjusting the domain weights according to a regression status of the different domains.

Abstract: The present invention relates to novel use of koumine, an alkaloid monomer from Gelsemium elegans Benth., or a pharmaceutically acceptable salt thereof, and more particularly to use of koumine or its pharmaceutically acceptable salt in the preparation of a medicament for the treatment of inflammatory bowel disease.

Abstract: A hydrogen bond induced high stability porous covalent organic gel material and a preparation method thereof are provided. The method comprises: dissolving tetrakis(4-carboxyphenyl)silane in methanol to obtain solution A; adding concentrated sulfuric acid to the solution A, then water-bath heating and re-flowing to obtain a solution B; evaporating the solution B, dissolving remaining powder with an ethyl acetate, washing and drying, filtering and then evaporating the filtrate until crystallization to obtain a colorless solid C; dissolving the colorless solid C and hydrazine hydrate in methanol, water-bath re-fluxing, filtering and collecting white powder, washing and drying to obtain a white solid D; adding the white solid D and 1,4-Phthalaldehyde to N,N-dimethyl-formamide, adding trifluoroacetic acid, and then getting the desired material.

Abstract: One or more fields can be divided into portions, and different sets of interventions can be performed during the course of planting and growing a crop in each field portion. The effect of each set of interventions on the crop outcome of the corresponding field portion can be determined using a best performing model of a set of predictive models applied to the set of interventions. The selected model can be used to determine an effect of each intervention on crop outcome of a field portion by estimating a portion of the crop outcome attributable to the intervention. A machine-learned model can be trained based on the effect each intervention has on crop outcome determined by the selected model, and the machine-learned model can be applied to subsequent fields and crops to predict effects that various interventions can have on subsequent crop outcomes.

Abstract: An augmentation and feature caching subsystem is described for training AI/ML models. In one particular aspect, a method is provided that includes receiving data comprising training examples, one or more augmentation configuration hyperparameters and one or more feature extraction configuration hyperparameters; generating a first key based on one of the training examples and the one or more augmentation configuration hyperparameters; searching a first key-value storage based on the first key; obtaining one or more augmentations based on the search of the first key-value storage; applying the obtained one or more augmentations to the training examples to result in augmented training examples; generating a second key based on one of the augmented training examples and the one or more feature extraction configuration hyperparameters; searching a second key-value storage based on the second key; obtaining one or more features based on the search of the second key-value storage.

Abstract: The present disclosure relates to a method and a system for determining a well spacing for in-situ leaching mining of a high-permeability sandstone-type uranium/copper deposit.

Abstract: The present disclosure provides a method and apparatus for quickly establishing communication, a device, and a storage medium. The method may include: in a non-standalone formed by a first standard network and a second standard network, a terminal determining whether a communication request to be sent to the first standard network is currently present; in a case where it is determined that the communication request to be sent to the first standard network is currently present, the terminal sending the communication request to the first standard network by using either of a current uplink resource of the first standard network and a current uplink resource of the second standard network; and after the communication request is sent to the first standard network, the terminal receiving, through the first standard network, a communication response sent by the first standard network in response to the communication request.

Abstract: An infusion liquid heating and flow-velocity-monitoring system for clinical use is provided. The system includes a dynamic heating module, a measurement and analysis module for average velocity in infusion tubes, a velocity adjustment module, an alarming and automatic clipping module, a mobile-phone-computer remote monitoring module, an operable shared module, and a micro control module. A method of determining average velocity on a cross section of a tube according to a solution to Fourier series in a dynamic thermal dispersion equation of a steady flow in the tube is provided. Moreover, functions including monitoring infusion rate, automatic heating, monitoring whether infusion fluid in the infusion tube is empty, and automatic clipping and closing of the tube for preventing blood backflow, etc. are achieved. Lastly, specific applications developed which is connected with mobile phones by 5G network, Bluetooth and WIFI are used to display status of the infusion fluid.

Abstract: Generally, the present disclosure is directed to user interface understanding. More particularly, the present disclosure relates to training and utilization of machine-learned models for user interface prediction and/or generation. A machine-learned interface prediction model can be pre-trained using a variety of pre-training tasks for eventual downstream task training and utilization (e.g., interface prediction, interface generation, etc.).

Abstract: A memory access method and device are provided for a Non-Uniform Memory Access (NUMA) system including nodes, each node configured to run a predetermined program and including a preset memory allocated for the predetermined program, where a same global physical address corresponds to each of the preset memories of each of the nodes, respectively, the method including: performing an operation on a first preset memory local to a first node of the nodes when the predetermined program is running on the first node; and synchronizing the operation performed on the first preset memory to each preset memory local to each of the nodes, respectively, where each of the nodes is configured to access its local preset memory using the same global physical address when running the predetermined program on the node.

Abstract: Novel techniques are described for data augmentation using a two-stage entity-aware augmentation to improve model robustness to entity value changes for intent prediction.

Abstract: Novel techniques are described for positive entity-aware augmentation using a two-stage augmentation to improve the stability of the model to entity value changes for intent prediction. In one particular aspect, a method is provided that includes accessing a first set of training data for an intent prediction model, the first set of training data comprising utterances and intent labels; applying one or more positive data augmentation techniques to the first set of training data, depending on the tuning requirements for hyper-parameters, to result in a second set of training data, where the positive data augmentation techniques comprise Entity-Aware (“EA”) technique and a two-stage augmentation technique; combining the first set of training data and the second set of training data to generate expanded training data; and training the intent prediction model using the expanded training data.