Abstract: In order to address a conventional problem that there is no information processing apparatus for simulating processing in the brain, an information processing apparatus is configured such that one or more pieces of feature information are transferred between somas, each soma may fire using one or more pieces of accepted feature information, a firing pattern, which is a pattern of firing of one or more somas, is acquired, and output information corresponding to the firing pattern is acquired and output. Accordingly, it is possible to realize information processing for simulating processing in the brain. Also, it is possible to realize information processing for simulating processing in the brain, such as growth processing, apoptosis processing, and learning processing of elements such as somas.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media for training a machine learning model to perform multiple machine learning tasks from multiple machine learning domains. One system includes a machine learning model that includes multiple input modality neural networks corresponding to respective different modalities and being configured to map received data inputs of the corresponding modality to mapped data inputs from a unified representation space; an encoder neural network configured to process mapped data inputs from the unified representation space to generate respective encoder data outputs; a decoder neural network configured to process encoder data outputs to generate respective decoder data outputs from the unified representation space; and multiple output modality neural networks corresponding to respective different modalities and being configured to map decoder data outputs to data outputs of the corresponding modality.

Abstract: An object detection device and an object detection method based on a neural network are provided. The object detection method includes: receiving an input image and identifying an object in the input image according to an improved YOLO-V2 neural network. The improved YOLO-V2 neural network includes a residual block, a third convolution layer, and a fourth convolution layer. A first input of the residual block is connected to a first convolution layer of the improved YOLO-V2 neural network, and an output of the residual block is connected to a second convolution layer of the improved YOLO-V2 neural network. Here, the residual block is configured to transmit, to the second convolution layer, a summation result corresponding to the first convolution layer. The third convolution layer and the fourth convolution layer are generated by decomposing a convolution layer of an original YOLO-V2 neural network.

Abstract: A signal processing device includes a receiver that receives a plurality of playback signal sequence obtained by digitizing a plurality of reading results by a plurality of A/D converter, the plurality of reading results being obtained by reading data by a plurality of reading elements from a magnetic tape and a plurality of equalizers that perform waveform equalization of the plurality of playback signal sequence. The plurality of equalizers perform the waveform equalization by using a plurality of non-linear filters that have been learned to reduce distortion that occurs non-linearly in the plurality of playback signal sequence according to a condition under an environment in which the data is read from the magnetic tape. The plurality of non-linear filters being optimized to a suitable characteristic for the plurality of reading elements by optimization based on the plurality of reading results.

Abstract: Provided is an intelligent analysis system for inner detecting magnetic flux leakage (MFL) data in pipelines, including a complete data set building module, a discovery module, a quantization module and a solution module, wherein: a complete data set building method is adopted in the complete data set building module to obtain a complete magnetic flux leakage data set; a pipeline connecting component discovery method is adopted in the discovery module to obtain the precise position of a weld; an anomaly candidate region search and identification method is adopted in the discovery model to find out magnetic flux leakage signals with defects; a defect quantization method based on a random forest is adopted in the quantization module to obtain a defect size; and a pipeline solution based on an improved ASME B31G standard is adopted in the solution module to output an evaluation result.

Abstract: Systems, devices, and methods related to a Deep Learning Accelerator and memory are described. For example, a removable media (e.g., a memory card, or a USB drive) may be configured to execute instructions with matrix operands and configured with: an interface to receive a video stream; and random access memory to buffer a portion of the video stream as an input to an Artificial Neural Network and to store instructions executable by the Deep Learning Accelerator and matrices of the Artificial Neural Network. Such a removable media can be used to replace an existing removable media used in a surveillance camera to record video or images. The Deep Learning Accelerator can execute the instructions to generate analytics of the buffer portion using the Artificial Neural Network, enabling the surveillance camera that is upgraded via the use of the removable media to provide intelligent services based on the analytics.

Abstract: The present disclosure sets forth a marginal distillation approach to obtaining a unified name-entity recognition (NER) student model from a plurality of pre-trained teacher NER models with different tag sets. Knowledge from the teacher models is distilled into a student model without requiring access to the annotated training data used to train the teacher models. In particular, the system receives a tag hierarchy that combines the different teacher tag sets. The teacher models and the student model are applied to a set of input data sequence to obtain tag predictions for each of the models. A distillation loss is computed between the student and each of the teacher models. If teacher's predictions are less fine-grained than the student's with respect to a node in the tag hierarchy, the student's more fine-grained predictions for the node are marginalized in computing the distillation loss.

Abstract: An information processing apparatus comprising a traffic line information generation unit configured to generate, based on information provided from an image capturing apparatus arranged in a store, traffic line information in the store of a customer who visits the store, an acquisition unit configured to acquire purchased product information associated with a product purchased by the customer in the store, a purchase information generation unit configured to generate purchase information based on the purchased product information by associating the traffic line information and the purchased product with the customer, and a display information generation unit configured to generate, based on the purchase information, display information for displaying information about one of a customer and a product in association with a time taken to select the purchased product and a selection order of the purchased product.

Abstract: A method includes receiving, with a computing device, a configuration file and a client request to apply a machine learning model to a set of data from a sensor. The method includes performing, with the computing device, preprocessing on the set of data from the sensor based on the configuration file to generate preprocessed data. The method includes sending, with the computing device, a call to a model server to apply the machine learning model to the preprocessed data.

Abstract: A distributed deep learning network may prevent an attacker from reconstructing raw data from activation outputs of an intermediate layer of the network. To achieve this, the loss function of the network may tend to reduce distance correlation between raw data and the activation outputs. For instance, the loss function may be the sum of two terms, where the first term is weighted distance correlation between raw data and activation outputs of a split layer of the network, and the second term is weighted categorical cross entropy of actual labels and label predictions. Distance correlation with the entire raw data may be minimized. Alternatively, distance correlation with only with certain features of the raw data may be minimized, in order to ensure attribute-level privacy. In some cases, a client computer calculates decorrelated representations of raw data before sharing information about the data with external computers.

Abstract: A classification apparatus is configured to perform a classification using a neural network with at least one hidden layer and an output layer, wherein the classification apparatus comprises a coarse training unit configured to train the neural network on a subset of neurons of a last hidden layer and a set of neurons of the output layer; and a fine training unit configured to train the neural network on a set of the last hidden layer and a subset of neurons of the output layer. By executing the classification apparatus, training of a classification model can be improved by reducing the computational burden of the classification, speeding up the training time of a classification model, and speeding up the inference time during application of the classification model.

Abstract: A database including various datasets and metadata associated with each respective dataset is provided. These datasets were used to train predictive models. The database stores a performance value associated with the model trained with each dataset. When provided with a new dataset, a server can determine various metadata for the new dataset. Using the metadata, the server can search the database and retrieve datasets which have similar metadata values. The server can narrow the search based on the performance value associated with the dataset. Based on the retrieved datasets, the server can recommend at least one sampling technique. The sampling technique can be determined based on the one or more sampling techniques that were used in association with the retrieved datasets.

Abstract: Disclosed are a system, apparatus and techniques for evaluating a dataset to confirm that the data in the dataset satisfies a data quality metric. A machine learning engine or the like may evaluate text strings within the dataset may be of arbitrary length and encoded according to an encoding standard. Data vectors of a preset length may be generated from the evaluated text strings using various techniques. Each data vector may be representative of the content of the text string and a category may be assigned to the respective data vector. The category assigned to each data vectors may be evaluated with respect to other data vectors in the dataset to determine compliance with a quality metric. In the case that a number of data vectors fail to meet a predetermined quality metric, an alert may be generated to mitigate any system errors that may result from unsatisfactory data quality.

Abstract: As described herein, a machine learning system, method, and computer program are provided to predict which resident of a residential space is watching television for content targeting purposes. In use, a login to a television service on a television device in a residential space is detected. Additionally, information defining a plurality of residents of the residential space is identified. Further, a profile determined for the login is identified, where the profile is associated with a particular resident of the plurality of residents or a particular resident group of the plurality of residents. Still yet, the profile and the information defining the plurality of residents of the residential space is input to a machine learning model to predict one or more residents of the plurality of residents that is consuming the television service on the television device.

Abstract: A method, system and computer-program product for identifying neural network inputs for a neural network that may have been incorrectly processed by the neural network. A set of activation values (of a subset of neurons of a single layer) associated with a neural network input is obtained. A neural network output associated with the neural network input is also obtained. A determination is made as to whether a first and second neural network input share similar sets of activation values, but dissimilar neural network outputs or vice versa. In this way a prediction can be made as to whether one of the first and second neural network inputs has been incorrectly processed by the neural network.

Abstract: Techniques for building and managing data models are provided. According to certain aspects, systems and methods may enable a user to input parameters associated with building one or more data models, including parameters associated with sampling, binning, and other factors. The systems and methods may automatically generate program code that corresponds to the inputted parameters and display the program code for review by the user. The systems and methods may build the data models and generate charts and plots depicting aspects of the data models. Additionally, the systems and methods may combine data models and select champion data models.

Abstract: A data clustering device includes an input configured to receive a plurality of data points encoded in at least one signal and a hardware logic circuit configured to extract one or more features of the one or more data points from the at least one signal, create or update, based on the one or more features, one or more data clusters representing one or more of the data points, and encode at least one of the one or more data clusters in at least one output signal. The device further includes an output configured to provide the at least one output signal, for instance, to a processor, such as a processor for controlling a controlled system. The device can be further configured to split or merge the data cluster(s) based on a statistical distribution of the one or more data points in the respective data cluster.

Abstract: A system and method for controlling a nodal network. The method includes estimating an effect on the objective caused by the existence or non-existence of a direct connection between a pair of nodes and changing a structure of the nodal network based at least in part on the estimate of the effect. A nodal network includes a strict partially ordered set, a weighted directed acyclic graph, an artificial neural network, and/or a layered feed-forward neural network.

Abstract: Anomalies in a data set may be difficult to detect when individual items are not gross outliers from a population average. Disclosed is an anomaly detector that includes neural networks such as an auto-encoder and a discriminator. The auto-encoder and the discriminator may be trained on a training set that does not include anomalies. During training, an auto-encoder generates an internal representation from the training set, and reconstructs the training set from the internal representation. The training continues until data loss in the reconstructed training set is below a configurable threshold. The discriminator may be trained until the internal representation is constrained to a multivariable unit normal. Once trained, the auto-encoder and discriminator identify anomalies in the evaluation set. The identified anomalies in an evaluation set may be linked to transaction, security breach or population trends, but broadly, disclosed techniques can be used to identify anomalies in any suitable population.

Abstract: Some examples relate generally to computer architecture software for information security and, in some more particular aspects, to machine learning based on changes in snapshot metadata for anomaly and ransomware detection in a file system.

Abstract: Disclosed are a method, a device, a system and/or a manufacture of iterative development and/or scalable deployment of a spreadsheet-based formula algorithm. In one embodiment, a system for scalable application of a data model defined in a spreadsheet to a dataset includes a translation server and an execution server. The translation server receives a spreadsheet file including a formula algorithm. The formula algorithm includes one or more spreadsheet formulas stored in cells. The translation server generates an extrapolated algorithm expressed in a programming language based on the formula algorithm. The execution server receives the extrapolated algorithm and the dataset and verifies calculation independence when applied to a data entry. The extrapolated algorithm is applied against the dataset. An iteration engine may continuously reapply the extrapolated algorithm to update an output data as the dataset evolves and/or receive an update to the formula algorithm and reapply the extrapolated algorithm.

Abstract: In one embodiment, L dimensional images are trained, mapped, and aligned to an M dimensional topology to obtain azimuthal angles. The aligned L dimensional images are then trained and mapped to an N dimensional topology to obtain 2N vertex classifications. The azimuthal angles and the 2N vertex classifications are used to map L dimensional images into 0 dimensional images.

Abstract: A system and method determine a classification by simulating a human user. The system and method translate an input segment such as speech into an output segment such as text and represents the frequency of words and phrases in the textual segment as an input vector. The system and method process the input vector and generate a plurality of intents and a plurality of sub-entities. The processing of multiple intents and sub-entities generates a second multiple of intents and sub-entities that represent a species classification. The system and method select an instance of an evolutionary model as a result of the recognition of one or more predefined semantically relevant words and phrases detected in the input vector.

Abstract: One embodiment can provide a system for detecting anomaly for high-dimensional sensor data associated with one or more machines. During operation, the system can obtain sensor data from a set of sensor associated with one or machines, apply data exploration techniques on the sensor data to automatically process sensor data to identify a subset of feature sensors from the available set of feature sensors, apply an unsupervised machine-learning technique to the identified subset of feature sensors and the target sensor to learn a set of pair-wise univariate models, and determine whether and how an anomaly occurs in the operation of the one or more machines based on the set of pair-wise univariate models.

Abstract: A non-transitory computer-readable recording medium stores therein a predicting program that causes a computer to execute: receiving input data to be subjected to prediction; and generating, from training data sets each having explanatory variables and an objective variable, a prediction result, by using a hypothesis set and respective weights of hypotheses included in the hypothesis set, the hypotheses each being formed of a combination of the explanatory variables, classifying any of the training data sets and satisfying a specific condition, the weights being learnt based on whether each of the hypotheses holds true for each of the training data sets. The generating includes determining a value of a variable included in a pseudo-Boolean function such that a probability satisfies a predetermined standard, the probability being a probability that the prediction result satisfies the specific condition, the pseudo-Boolean function including variables corresponding to the explanatory variables.

Abstract: A unique implementation of a machine learning application for suggesting actions for a user to undertake is described herein. The application transforms a history of user behavior into a set of models that represent user actions given a set of parameters. These models are then used to suggest that users in a payments or banking environment take certain actions based on their history. The models are created using the DensiCube, random forest or k-means algorithms.

Abstract: Provided herein are systems and methods for applying adaptive classes thresholds to enhance object detection Machine Learning (ML) models by receiving a plurality of labeled feature vectors extracted from a plurality of images associated with a plurality of objects, one or more subsets of the plurality of feature vectors are associated with respective object(s) and labeled accordingly, computing an adaptive threshold for each object in a plurality of iterations, each iteration comprising: (1) computing deviation of a respective feature vector of the subset from an aggregated feature vector, (2) computing, in case the deviation is within a predefined value, a threshold enclosing the respective feature vector, and (3) adjusting the adaptive threshold to enclose the threshold of the respective feature vector and outputting the adaptive threshold(s) for classifying unlabeled feature vectors to class(s) of respective object(s) associated with the adaptive threshold(s) in which the unlabeled feature vectors fall.

Abstract: In implementations of the subject matter described herein, a solution for object detection is proposed. First, a feature(s) is extracted from an image and used to identify a candidate object region in the image. Then another feature(s) is extracted from the identified candidate object region. Based on the features extracted in these two stages, a target object region in the image and a confidence for the target object region are determined. In this way, the features that characterize the image from the whole scale and a local scale are both taken into consideration in object recognition, thereby improving accuracy of the object detection.

Abstract: A computer-implemented user profiling system includes a human communication retrieval component which, for an entity employing a business process management system, captures communication data in response to a given business process being implemented by the business process management system. A human task monitoring and contextual analysis component captures user behavior information associated with the business process. A profile analysis engine is also included, which receives the user behavior information and communication data and updates a user profile corresponding to at least one of the users, based on the user behavior information and communication data.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training a policy neural network used to select actions to be performed by a reinforcement learning agent interacting with an environment. In one aspect, a method includes obtaining path data defining a path through the environment traversed by the agent. A consistency error is determined for the path from a combined reward, first and last soft-max state values, and a path likelihood. A value update for the current values of the policy neural network parameters is determined from at least the consistency error. The value update is used to adjust the current values of the policy neural network parameters.

Abstract: A method for training a computerized mechanical device, comprising: receiving data documenting actions of an actuator performing a task in a plurality of iterations; calculating using the data a neural network dataset and used for performing the task; gathering in a plurality of reward iterations a plurality of scores given by an instructor to a plurality of states, each comprising at least one sensor value, while a robotic actuator performs the task according to the neural network; calculating using the plurality of scores a reward dataset used for computing a reward function; updating at least some of the neural network's plurality of parameters by receiving in each of a plurality of policy iterations a reward value computed by applying the reward function to another state comprising at least one sensor value, while the robotic actuator performs the task according to the neural network; and outputting the updated neural network.

Abstract: Disclosed are a method and apparatus for SAR image data enhancement, and a storage medium. The method includes: processing an SAR target image by electromagnetic simulation to acquire an SAR electromagnetic simulation image; and processing the SAR electromagnetic simulation image and the SAR target image by a generative adversarial network to obtain a set of virtual samples of the SAR target image.

Abstract: The present disclosure relates to a method, an apparatus and a system for controlling a robot, and a storage medium. The method uses a neural network connected to an external memory to conduct the controlling of the robot, and comprises: inputting input data into the learned neural network to obtain output data, wherein said input data comprises an image about an object, said output data comprises control data about said robot; and establishing an association between part or all of the information generated by said neural network during the calculation and said input data and/or said output data, wherein said part or all of the information represents a feature of said object related to said control data. Thus, the user can grasp the calculation process of the neural network.

Abstract: The exemplary embodiments disclose a method, a computer program product, and a computer system for mitigating the risks associated with handling items. The exemplary embodiments may include collecting data relating to one or more items, extracting one or more features from the collected data, determining one or more hazards based on the extracted one or more features and one or more models, and displaying the one or more hazards within an augmented reality device worn by a user.

Abstract: Apparatus and associated methods relate to providing a machine learning methodology that uses the machine learning's own failure experiences to optimize future solution search and provide self-guided information (e.g., the dependency and independency among various adaptation behavior) to predict a receiver's equalization adaptations. In an illustrative example, a method may include performing a first training on a first neural network model and determining whether all of the equalization parameters are tracked. If not all of the equalization parameters are tracked under the first training, then, a second training on a cascaded model may be performed. The cascaded model may include the first neural network model, and training data of the second training may include successful learning experiences and data of the first neural network model. The prediction accuracy of the trained model may be advantageously kept while having a low demand for training data.

Abstract: Techniques for model reutilization with heterogeneous sensor stacks via sensor data auto-normalization are described. A normalization model can be trained and utilized to normalize sensor data generated by a first type of sensor stack so that it can be used with an existing machine learning model that was trained using data from another type or types of sensor stacks having different characteristics. A sensor data can be generated by the sensor stack and provided as an input to the normalization model to yield normalized sensor data. The normalized sensor data can be provided as input to the existing model to generate accurate results despite the sensor stack having different characteristics than the one(s) used to train the machine learning model.

Abstract: This disclosure relates to improved techniques for performing human parsing functions using neural network architectures. The neural network architecture can model human objects in images using a hierarchal graph of interconnected nodes that correspond to anatomical features at various levels. Multi-level inference information can be generated for each of the nodes using separate inference processes. The multi-level inference information for each node can be combined or fused to generate final predictions for each of the nodes. Parsing results may be generated based on the final predictions.

Abstract: A system and method for attribute discovery for operation objects from operation data includes segmenting a name of each of a plurality of operation objects based on one or more special characters used in the name of each operation object. A similarity comparison of the operation objects is performed by extracting common subsequences from substrings in operation data in a same log as a target object, and a string similarity is computed of the extracted common subsequences. Numerical attributes are determined by calculating statistical metrics for fields in the log, and additional information of the operation objects is discovered based on the determined numerical attributes.

Abstract: This disclosure relates to methods and systems for simulation of electricity value ecosystem using agent based modeling approach. State-of-the-art methods utilize simulation tools to support decision making that do not model agents own behaviour and its response to other agents based on an interaction, thereby unable to analyse complex interactions in the electricity value ecosystem. The present disclosure provides a generalized integrated simulation platform which provides dynamic configurability to simulate a plurality of user requirements associated with the electricity value eco-system using a causal diagram which is further used to identify a plurality of agents. Further, a plurality of a plurality of models and processes for the plurality of agents are determined or generated based on their availability in a repository. The causal diagram is refined in accordance with one or more constraints which supports in making a better and informed decision considering changing dynamics of the plurality of agents.

Abstract: Methods and systems for communicating with a server of a cloud services system used to interface with vehicles are provided. One method includes receiving, by the server, a request from electronics of a vehicle to access a profile for a user account. The request identifies user information for a user to use the vehicle. The method includes processing, by the server, at least part of the user information to verify the user against data associated with the user account. The profile has a plurality of settings of the user for the vehicle, and at least part of the plurality of settings for the profile being stored on storage accessible to the cloud services system. The method includes transferring, by the server, upon verification of the user information, one or more settings of the plurality of settings to the vehicle.

Abstract: A virtual agent with a dialogue management system and a method of training the dialogue management system is disclosed. The dialogue management system is trained using a deep reinforcement learning process. Training involves obtaining or simulating training dialogue data. During the training process, actions for the dialogue management system are selected using a Deep Q Network to process observations. The Deep Q Network is updated using a target function that includes a reward. The reward may be generated by considering one or more of the following metrics: task completion percentage, dialogue length, sentiment analysis of the user's response, emotional analysis of the user's state, explicit user feedback, and assessed quality of the action. The set of actions that the dialogue management system can take at any time may be limited by an action screener that predicts the subset of actions that the agent should consider for a given state of the system.

Abstract: A system for end to end prediction of lane detection uncertainty includes a sensor device of a host vehicle generating data related to a road surface and a navigation controller including a computerized processor operable to monitor an input image from the sensor device, utilize a convolutional neural network to analyze the input image and output a lane prediction and a lane uncertainty prediction, and generate a commanded navigation plot based upon the lane prediction and the lane uncertainty prediction. The convolutional neural network is initially trained using a per point association and error calculation, including associating a selected ground truth lane to a selected set of data points related to a predicted lane and then associating at least one point of the selected ground truth lane to a corresponding data point from the selected set of data points related to the predicted lane.

Abstract: Techniques are provided herein for creating well-balanced computer-based reasoning systems and using those to control systems. The techniques include receiving a request to determine whether to use one or more particular data elements, features, cases, etc. in a computer-based reasoning model (e.g., as data elements, cases or features are being added, or as part of pruning existing features or cases). Conviction measures are determined and inclusivity conditions are tested. The result of comparing the conviction measure can be used to determine whether to include or exclude the feature, case, etc. in the model and/or whether there are anomalies in the model. A controllable system may then be controlled using the computer-based reasoning model.

Abstract: A data stream processing system can receive a stream of data and display a portion of the stream to a user. The displayed streaming data can change over time as additional data is received as part of the stream. The data stream processing system can extract one or more field values rom data in the stream and generate filters based on the extracted information. The generated filters can be displayed to a user, and in response to an interaction with a generated filter, the data stream processing system can apply the selected filter to data in the data stream.

Abstract: System, methods, and other embodiments described herein relate to semi-supervised training of a depth model for monocular depth estimation. In one embodiment, a method includes training the depth model according to a first stage that is self-supervised and that includes using first training data that comprises pairs of training images. Respective ones of the pairs including separate frames depicting a scene of a monocular video. The method includes training the depth model according to a second stage that is weakly supervised and that includes using second training data to produce depth maps according to the depth model. The second training data comprising individual images with corresponding sparse depth data. The second training data providing for updating the depth model according to second stage loss values that are based, at least in part, on the depth maps and the depth data.

Abstract: Embodiments for training a neural network are provided. A neural network is divided into a first block and a second block, and the parameters in the first block and second block are trained in parallel. To train the parameters, a gradient from a gradient mini-batch included in training data is generated. A curvature-vector product from a curvature mini-batch included in the training data is also generated. The gradient and the curvature-vector product generate a conjugate gradient. The conjugate gradient is used to determine a change in parameters in the first block in parallel with a change in parameters in the second block. The curvature matrix in the curvature-vector product includes zero values when the terms correspond to parameters from different blocks.

Abstract: A method for digital image segmentation is provided. The method comprises training a neural network for image segmentation with a labeled training dataset from a first domain, wherein a subset of nodes in the neural net are dropped out during training. The neural network receives image data from a second, different domain. A vector of N values that sum to 1 is calculated for each image element, wherein each value represents an image segmentation class. A label is assigned to each image element according to the class with the highest value in the vector. Multiple inferences are performed with active dropout layers for each image element, and an uncertainty value is generated for each image element. The label of any image element with an uncertainty value above a predefined threshold is replaced with a new label corresponding to the class with the next highest value.

Abstract: Individual computer diagnostic support (CDS) systems are coupled to a ‘global’ CDS system, each of the CDS systems using the same learning system or the same learning system technique. Training and testing cases from each of the individual CDS systems are provided to the global CDS system, and the global CDS system uses these training cases to produce learning system parameters based on the training cases. Having more training cases than any of the individual CDS systems, the parameters provided by the global CDS system offer a higher quality diagnostic output than any of the individual CDS systems. The learning system parameters at the global CDS system may be provided to each of the individual CDS systems, to update the parameters of the individual CDS systems' learning system. The global CDS may also refine and/or adjust the structure of the embodied learning systems.

Abstract: Provided is a computer system for estimating an absolute position of a photographed object by just photographing the object with a camera, a position estimation method, and a program. The computer system acquires an image obtained by photographing an object, acquires three-dimensional position data of a camera which photographed the object, and estimates an absolute position of the object on the basis of the three-dimensional position data of the camera. Further, the computer system enables the camera to be tilted a specified angle in a direction of the object, and estimates the absolute position of the object on the basis of the three-dimensional position data of the camera and the tilted specified angle. Moreover, the computer system stores the position of the object and an altitude at the position in association with each other, and estimates an altitude associated with the estimated position of the object.

Abstract: A computerized-system and method for generating a reduced-size superior labeled training-dataset for a high-accuracy machine-learning-classification model for extreme class imbalance by: (a) retrieving minority and majority class instances to mark them as related to an initial dataset; (b) retrieving a sample of majority instances; (c) selecting an instance to operate a clustering classification model on it and the instances marked as related to the initial dataset to yield clusters; (d) operating a learner model to: (i) measure each instance in the yielded clusters according to a differentiability and an indicativeness estimators; (ii) mark measured instances as related to an intermediate training dataset according to the differentiability and the indicativeness estimators; (e) repeating until a preconfigured condition is met; (f) applying a variation estimator on all marked instances as related to an intermediate training dataset to select most distant instances; and (g) marking the instances as related to