Abstract: A computer-implemented method for learning a model to predict movements of a person in bed is presented. The method includes receiving first data from a plurality of first sensors installed on a bed, receiving second data from a plurality of second sensors installed on the person, and learning a model to predict the second data based on the first data by assuming a sensing range of motion intensity by the plurality of first sensors is greater than a sensing range of motion intensity by the plurality of second sensors.

Abstract: A computer implemented method of generating new chemical compounds is provided. The method includes preparing a data-driven substructure feature vector for each of a plurality of chemical compounds for which a chemical or physical property is known. The method further includes preparing a predefined component feature vector, creating a regression model to predict a target value for the chemical or physical property, and performing a search algorithm to identify substructure features that affect the target value for the chemical or physical property. The method further includes generating a candidate structure having the target value for the chemical or physical property, and synthesizing the candidate structure.

Abstract: A non-transitory computer readable storage medium having instructions embodied therewith, the instructions executable by a processor or programmable circuitry to cause the processor or programmable circuitry to perform operations including obtaining training data including a sample value of one or more input features of an item and a sample value of an output feature representing demand for the item, and training, based on the training data, an estimation model that estimates a new value of the output feature for the item based on new values of the one or more input features. The one or more input features may include a relative price of the item relative to prices of a plurality of items.

Abstract: Cooperative neural networks reinforcement learning may be performed by obtaining an action and observation sequence, inputting each time frame of the action and observation sequence sequentially into a first neural network including a plurality of first parameters and a second neural network including a plurality of second parameters, approximating an action-value function using the first neural network, and updating the plurality of second parameters to approximate a policy of actions by using updated first parameters.

Abstract: A battery controller and method for controlling a battery include generating a battery capacity prediction model that characterizes a battery capacity decay rate. Future battery capacity for a battery under control is predicted based on the battery capacity prediction model and a present value of the battery capacity. One or more operational parameters of the battery under control is controlled based on the predicted future battery capacity.

Abstract: A method providing an analytical technique introducing label information into an anomaly detection model. Effective utilization of label information is based on introducing the degree of similarity between samples. Assuming, for example, there is a degree of similarity between normally labeled samples and no similarity between normally labeled and abnormally labeled samples. Also each sensor value is generated by the linear sum of a latent variable and a coefficient vector specific to each sensor. However, the magnitude of observation noise is formulated to vary according to the label information for the sensor values, and set so that normal label?unlabeled?anomalously labeled. A graph Laplacian is created based on the degree of similarity between samples, and determines the optimal linear transformation matrix according to a gradient method. A optimal linear transformation matrix is used to calculate an anomaly score for each sensor in the test samples.

Abstract: A method and apparatus for data processing. The present invention provides a data processing apparatus that includes: a series acquisition section for acquiring a data series in which multiple pieces of data are arranged; a fragmentation section for fragmenting the data series to obtain multiple partial data series; a pattern extraction section for extracting multiple patterns of one or more pieces of data appearing in at least one of the multiple partial data series; and a generation section for generating a feature vector having element values, which vary according to whether to include each of the multiple patterns, for each of the multiple partial data series, respectively. There is also provided a method for data processing. The present invention allows for the generation of a feature vector from time-series data indicating a phenomenon the occurrence time of which is temporally irregular to detect features.

Abstract: Similarity of items can be estimated by using a method including generating a prediction model that predicts an indicator of a target based on one or more attributes for items, by estimating a weight set, among weight sets, for each of the items, and estimating a similarity among the items for the target based on the weight sets of the prediction model.

Abstract: Similarity of items can be estimated by using a method including generating a prediction model that predicts an indicator of a target based on one or more attributes for items, by estimating a weight set, among weight sets, for each of the items, and estimating a similarity among the items for the target based on the weight sets of the prediction model.

Abstract: A method is provided for rule creation that includes receiving (i) a MDP model with a set of states, a set of actions, and a set of transition probabilities, (ii) a policy that corresponds to rules for a rule engine, and (iii) a set of candidate states that can be added to the set of states. The method includes transforming the MDP model to include a reward function using an inverse reinforcement learning process on the MDP model and on the policy. The method includes finding a state from the candidate states, and generating a refined MDP model with the reward function by updating the transition probabilities related to the state. The method includes obtaining an optimal policy for the refined MDP model with the reward function, based on the reward policy, the state, and the updated probabilities. The method includes updating the rule engine based on the optimal policy.

Abstract: A method is provided for rule creation that includes receiving (i) a MDP model with a set of states, a set of actions, and a set of transition probabilities, (ii) a policy that corresponds to rules for a rule engine, and (iii) a set of candidate states that can be added to the set of states. The method includes transforming the MDP model to include a reward function using an inverse reinforcement learning process on the MDP model and on the policy. The method includes finding a state from the candidate states, and generating a refined MDP model with the reward function by updating the transition probabilities related to the state. The method includes obtaining an optimal policy for the refined MDP model with the reward function, based on the reward policy, the state, and the updated probabilities. The method includes updating the rule engine based on the optimal policy.

Abstract: Variation of attribute values of agents are estimated by performing a method including obtaining a first matrix that represents sales records, each sales record including an item among items, a sales event among sales events, and an agent among agents. The first matrix is decomposed into a second matrix that represents first attribute values for each of the sales events, and a third matrix that represents the first attribute values for each of the items. The first attribute values are calculated for an agent among the agents, based on the second matrix, and a variation of the first attribute values for the agent are estimated as caused by a variation of sales records of each of the item among the items.

Abstract: Variation of attribute values of agents are estimated by performing a method including obtaining a first matrix that represents sales records, each sales record including an item among items, a sales event among sales events, and an agent among agents. The first matrix is decomposed into a second matrix that represents first attribute values for each of the sales events, and a third matrix that represents the first attribute values for each of the items. The first attribute values are calculated for an agent among the agents, based on the second matrix, and a variation of the first attribute values for the agent are estimated as caused by a variation of sales records of each of the item among the items.

Abstract: A battery controller and method for controlling a battery include generating a battery capacity prediction model that characterizes a battery capacity decay rate. Future battery capacity for a battery under control is predicted based on the battery capacity prediction model and a present value of the battery capacity. One or more operational parameters of the battery under control is controlled based on the predicted future battery capacity.

Abstract: Probability density ratios may be estimated by a method including obtaining a first sample set including a plurality of first samples and a second sample set including a plurality of second samples, wherein each of the first samples and the second samples is represented as a vector including a plurality of parameters, constructing at least one decision tree estimating a ratio of probability density p(x)/q(x) based on the first sample set and the second sample set, wherein p(x) is a probability density of the first samples corresponding to an input vector x and q(x) is a probability density of the second samples corresponding to the input vector x.

Abstract: Probability density ratios may be estimated by a method including obtaining a first sample set including a plurality of first samples and a second sample set including a plurality of second samples, wherein each of the first samples and the second samples is represented as a vector including a plurality of parameters, constructing at least one decision tree estimating a ratio of probability density p(x)/q(x) based on the first sample set and the second sample set, wherein p(x) is a probability density of the first samples corresponding to an input vector x and q(x) is a probability density of the second samples corresponding to the input vector x.

Abstract: A training method is provided. The training method includes clustering, by a processor, a plurality of items that each have an item attribute value, according to the item attribute value. The training method further includes generating, by the processor, for each item, a cluster attribute value corresponding to a cluster associated with the item. The training method also includes training, by the processor, an estimation model for estimating selection behavior of a target with respect to a choice set including two or more items, based on the cluster attribute value associated with each item included in the choice set, by using training data that includes a group of a choice set of items presented to the target and an item selected by the target from among the choice set.

Abstract: A training method is provided. The training method includes clustering, by a processor, a plurality of items that each have an item attribute value, according to the item attribute value. The training method further includes generating, by the processor, for each item, a cluster attribute value corresponding to a cluster associated with the item. The training method also includes training, by the processor, an estimation model for estimating selection behavior of a target with respect to a choice set including two or more items, based on the cluster attribute value associated with each item included in the choice set, by using training data that includes a group of a choice set of items presented to the target and an item selected by the target from among the choice set.

Abstract: A battery controller and method for controlling a battery include training parameters for a battery capacity prediction model based usage of similar batteries and capacity information for the respective similar batteries. The model characterizes a capacity decay rate. Future battery capacity is predicted for a battery under control based on the battery capacity prediction model and a present value of the battery capacity. One or more operational parameters of the battery under control is controlled based on the predicted future battery capacity.

Abstract: A method for selecting an action, includes reading, into a memory, a Partially Observed Markov Decision Process (POMDP) model, the POMDP model having top-k action IDs for each belief state, the top-k action IDs maximizing expected long-term cumulative rewards in each time-step, and k being an integer of two or more, in the execution-time process of the POMDP model, detecting a situation where an action identified by the best action ID among the top-k action IDs for a current belief state is unable to be selected due to a constraint, and selecting and executing an action identified by the second best action ID among the top-k action IDs for the current belief state in response to a detection of the situation. The top-k action IDs may be top-k alpha vectors, each of the top-k alpha vectors having an associated action, or identifiers of top-k actions associated with alpha vectors.