A construction machine includes a travel actuator, an attachment actuator, a storage, an information obtaining device, and a hardware processor configured to perform braking control of at least one of the travel actuator and the attachment actuator in response to determining that a dangerous situation is going to occur based on information obtained by the information obtaining device and information stored in a database in the storage.

A reinforcement learning (RL) based adaptive state observation model usable for implementing a brain machine interface (BMI) is proposed for decoding a brain signal to determine a movement action and controlling a machine to perform the movement action. In the model, the brain signal is processed by a neural network (NN) for applying a nonlinear mapping defined by NN weights to the brain signal to thereby yield a transformed brain signal. The NN learns the nonlinear mapping by RL, allowing the weights to be adaptively and continuously updated to follow nonlinearity and non-stationarity of the brain signal. The transformed brain signal is processed by a Kalman filter (KF) to yield a control signal for controlling the machine to perform the movement action, thereby utilizing the KF to provide smooth generation of the control signal while blocking adverse influence of nonlinearity and non-stationarity of the brain signal to the KF.

An approach for a non-factoid question answering framework across tasks and domains may be provided. The approach may include training a multi-task joint learning model in a general domain. The approach may also include initializing the multi-task joint learning model in a specific target domain. The approach may include tuning the joint learning model in the target domain. The approach may include determining which task of the multiple tasks is more difficult for the multi-task joint learning model to learn. The approach may also include dynamically adjusting the weights of the multi-task joint learning model, allowing the model to concentrate on learning the more difficult learning task.

A method for dynamically assessing slope safety includes the following steps: S1, carrying out geologic model generalization to the slope according to slope type, slope structure, stratum characteristics and a deformation failure mode to obtain a slope geologic model, creating a slope geometric model according to the slope geologic model, carrying out the subdivision of computational grid, and selecting a reasonable numerical simulation method, mechanical constitutive and initial boundary value conditions to form a computational model; and S2, adjusting stratum parameters, structural plane parameters and activating factor strength based on the computational model, carrying out a large amount of numerical simulation, summarizing results of the numerical simulation, normalizing input quantities and output quantities to establish machine learning samples. The method is able to dynamically adjust the geomechanical input parameters by using the monitoring data, making the prediction accuracy further higher, and can further achieve the real-time prediction.

Methods and systems classify and cluster invoice data. An invoice is obtained. A category vector is generated from an invoice string of the invoice with a dense layer of a machine learning model that includes an embedding layer, a neural network layer, and the dense layer. A suggestion is selected with a selection engine and in response to comparing the category vector to a set of clusters. The suggestion is presented.

A social recommendation method based on a multi-feature heterogeneous graph neural network is provided and includes: extracting attribute information of users and topics to code; processing user coding information and topic coding information through a multi-layer perceptron to obtain initial feature vectors of the users and the topics; establishing a heterogeneous graph by taking the users and the topics as nodes; inputting the heterogeneous graph into a heterogeneous graph neural network to perform information transmission in combination with an attention mechanism, and updating the feature vectors; and performing similarity calculation on the feature vectors of the users, and selecting the user and the topic with the highest similarity with the feature vector of the user for recommendation. Social information can be mined more comprehensively, features of users and interested topics of the users can be deeply fused, and recommendation accuracy and user experience can be improved.

A method and a system for generating an interpretable embedding that corresponds to a sequence of events is provided. The method includes: receiving information that corresponds to a sequence of events that respectively correspond to interactions between a customer and an organization; determining, for each respective event, a respective product associated with the organization and a respective channel via which the event has occurred; assigning a respective sentiment to each event; computing a respective weight for each event; aggregating the computed weights with respect to the products and the channels; and using the aggregated weights to generate the interpretable embedding for the customer. The interpretable embedding is then usable for generating targeted offers to the customer, handling complaints, and preventing subsequent complaints.

A method for predicting a yield of calcium in a calcium treatment process based on deep neural network as provided relates to a calcium treatment process of molten steel refining in the field of iron and steel metallurgy, and includes steps of: obtaining production and operation data information of each of charges and thereby constructing a dataset; training and testing a deep neural network based on constructed dataset to establish a prediction model; and based on the prediction model, predicting and calculating current yield of calcium by taking actual production and operation data information of each charge as input. The method can predict the yield of calcium in the calcium treatment process, is favorable for accurately controlling a calcium content of steel, can stably control the calcium treatment process, improve the calcium treatment effect, improve the product quality, and ensure the production stability.

Technologies are provided for assigning developers to source code issues using machine learning. A machine learning model can be generated based on multiple versions of source code objects (such as source code files, classes, modules, packages, etc.), such as those that are managed by a version control system. The versions of the source code objects can reflect changes that are made to the source code objects over time. Associations between developers and source code object versions can be analyzed and used to train the machine learning model. Patterns of similar changes to various source code objects can be detected and can also be used to train the machine learning model. When an issue is detected in a version of a source code object, the model can be used to identify a developer to assign to the issue. Feedback data regarding the developer assignment can be used to re-train the model.

A method of classifying an object according to an embodiment includes extracting a first feature by transforming rectangular coordinates of points included in the box of the object, obtained from a point cloud acquired using a LiDAR sensor, into complex coordinates and performing Fast Fourier Transform (FFT) on the complex coordinates, obtaining an average and a standard deviation as a second feature, the average and the standard deviation being parameters of a Gaussian model for the points included in the box of the object, and classifying the type of object based on at least one of the first feature or the second feature.